Prevalence, duration and factors associated with breastfeeding in French Guiana in 2024

Abstract

Introduction

Exclusive breastfeeding is essential for health throughout life. The World Health Organization (WHO) recommends exclusive breastfeeding for the first six months, as well as continued breastfeeding for up to two years or more in addition to an appropriate diet. In French Guiana, a French overseas territory characterized by cultural diversity, socioeconomic inequalities, and structural healthcare challenges, there was a lack of recent, representative data on breastfeeding practices. The ANJE Nutri study aimed to estimate the prevalence of breastfeeding practices and to identify factors associated with exclusive breastfeeding at four months (EBF-4), the national benchmark defined by the French National Nutrition and Health Program (PNNS).

Methods

This multicenter survey was conducted by telephone with mothers included in the Nutri Pou Ti’moun study approximatively 12 months after delivering in one of French Guiana’s three maternity wards. Multilingual interviewers collected data using a structured questionnaire designed to measure Infant and Young Child Feeding (IYCF) indicators. Modified Poisson regression was used to identify factors associated with EBF-4.

Results

Between March and October 2024, 382 mother-child pairs were included in the study. The majority of infants (91.6%) were ever breastfed, including 89.8% in the maternity stay, with a median breastfeeding duration of six months. Among children born in the last 12 months, 51.6% were ever exclusively breastfed, including 20.1% in the maternity stay, with a median breastfeeding duration of three months. The prevalence of EBF-4 was 22.3% [95% CI: 17.8–27.6]. In the maternity stay, 77.4% of breastfed infants received maternity supplement (MS), mainly infant formula. Professional breastfeeding support was reported by 69.6% of mothers during the maternity stay and by 40.9% after discharge; 22.8% wished they had received support but did not. In the multivariate analysis, EBF-4 was positively associated with younger maternal age (under 20 years) [PR: 4.2; 95% CI: 2.3–7.7] and higher education level [PR: 2.0; 95% CI: 1.2–3.4]. EBF-4 was negatively associated with being born in French Guiana (compared to women born in Haiti) [PR: 0.4; 95% CI: 0.2–0.8] and with the introduction of maternity supplements [PR: 0.3; 95% CI: 0.2–0.5].

Conclusion

Overall breastfeeding rates in French Guiana are encouraging, yet EBF-4 remains low. The early and frequent use of MS and disparities in access to breastfeeding support were associated with lower likelihood of sustaining exclusive breastfeeding and breastfeeding. These findings underscore the necessity of providing early, coordinated, and culturally appropriate support strategies that are tailored to the diverse maternal profiles in this territory, particularly to empower mothers to make informed choices about infant feeding.

Introduction

The first months of life are considered a critical period that has a lasting influence on a child’s future health. The literature clearly establishes the benefits of breastfeeding (BF) in terms of preventing infections, reducing infant mortality, supporting neurocognitive development and contributing to healthy growth [1, 2]. BF also benefits maternal health by reducing the incidence of breast and ovarian cancer, improving postpartum recovery, facilitating faster weight loss [3–6] and strengthening the bond between mother and child [7].

According to the United Nations International Children’s Emergency Fund (UNICEF), exclusive breastfeeding (EBF) under six months, without water or infant formula, is the most appropriate nutritional solution for infants. Failure to breastfeed exclusively increases the risk of death during this period by up to 14 times [8]. EBF is considered a veritable “vaccine” and is therefore at the heart of public health policies, particularly those of the World Health Organization (WHO), which recommends EBF under six months, followed by extended BF until the age of two years or more, alongside appropriate dietary diversification [8, 9]. These recommendations have been translated into key indicators for infant and young child feeding (IYCF), which aim to evaluate practices at the population level [10]. Despite persistent disparities, particularly in areas marked by socioeconomic inequalities, and the acknowledgement that enhanced breastfeeding support is required to mitigate health disparities, UNICEF and WHO highlight an overall increase in EBF rate over the past decade. These rates have risen from 38% to 48% in 2023 [9].

In France, the National Nutrition and Health Program (PNNS), launched in 2001, aims to promote EBF. In line with WHO recommendations, the PNNS has adapted to the French context and has emphasized that introducing solid foods between four and six months of age is optimal. Therefore, EBF for the first four months of life is considered optimal [11–13]. It also included expertise on nutrition in overseas territories to “identify differences and similarities” and “guide public policy” [3] in view of the considerable differences in living conditions, cultures, and access to food, healthcare and prevention. French Guiana has specific epidemiological and social characteristics that may influence BF practices [14]. The territory consists of a vast mosaic of cultures stemming from indigenous peoples as well as successive waves of migration (notably from Haiti, Brazil, Suriname, and Europe) [15, 16]. Each of these groups has distinct cultural norms, practices, and perceptions regarding infant feeding. From the health system perspective, French Guiana faces structural challenges, including unequal access to healthcare depending on geographical location, particularly in remote inland communities along the Maroni and Oyapock rivers [16]. Finally, there is a persistent challenge at the level of healthcare personnel, who experience high staff turnover and limited availability, particularly of certified lactation consultants (IBCLC). In French Guiana, where infant mortality and poverty are significantly higher, recent data on nutrition remains limited [17]. Previous studies have revealed a high prevalence of food insecurity (32.0%) and micronutrient deficiencies (81.0% with at least one deficiency) among pregnant women in French Guiana. This raises concerns about the potential impact on the nutritional status of their children [18, 19].

To supplement and build on these findings, the ANJE Nutri study was conducted in French Guiana in 2024. Its aimed to produce recent and representative data on BF practices and to estimate the rates of breastfeeding, exclusive breastfeeding and exclusive breastfeeding at 4 months (EBF-4), and identify the factors influencing these practices. This information will inform public health actions and policies. Our primary outcome (EBF-4) was guided by the PNNS framework, which sets 4 months as the national benchmark for EBF in France [7, 11–13]. This allows comparison with French surveillance data (Epifane survey [20]).

Methods

Study design

The ANJE Nutri study was a multicenter study, ancillary to the cross-sectional Nutri Pou Ti’moun survey, which focused on food insecurity among pregnant women. The study was conducted at all maternity units in French Guiana in 2023 [18]. The Nutri Pou Ti’moun survey was based on a systematic, stratified probability sample with proportional allocation.

Study population

The inclusion criteria corresponded to women included in the Nutri Pou Ti’moun survey [18] who did not object to participating in the ANJE Nutri study (or whose legal representative did not object for underage women).

In cases of multiple births, only the information on one of the newborns was retained: the one whose first name was closest to the letter “A”. Mothers of deceased children or those not responsible for their children were not included.

To achieve a margin of error of 5%, a confidence level of 99%, and an expected rate of EBF-4 of 13% (the national reference rate in the 2021 Epifane study [20], the study required a sample size of 290 participants. The participation rate (based on previous telephone surveys in French Guiana and elsewhere [21]) was estimated at 40%.

According to the latest available computerized pregnancy outcome register (RIGi), 8,495 births were recorded in French Guiana [22]. Most births (97.5%) took place in one of the three public maternity hospitals: 47.5% at the Cayenne Hospital Center (CHC), 40.5% at the Western French Guiana Hospital Center (CHOG) and 9.5% at the Kourou Hospital Center (CHK).

Data collection methods

Mothers were contacted by multilingual interviewers by telephone (landline or mobile) when their child was approximately 12 months old. An inclusion procedure in the form of a decision tree was established to guide the interviewer: three attempts were made to call, followed by a message if necessary. After obtaining consent, the interviewer asked the mother to answer a questionnaire over the phone for 20 to 30 min.

Data collection

1. Telephone questionnaire: information on BF practices (type, duration, reasons for stopping).

2. Medical data: on women (behavior, medical history, gynecological and obstetric history, height and weight) and newborns (height, weight, prematurity).

The questionnaire was developed by the research team based on standardized indicators for IYCF, as defined by PNNS and used in the French Epifane study [20], in order to facilitate data comparison. It was also developed in accordance with the WHO’s guide on indicators, definitions and measurement methods, for assessing IYCF practices [10].

Before enrolment in the study began, multilingual interviewers received academic training on the basic principles of IYCF (French and international recommendations, definitions of key indicators and the timeline of feeding during the first year). This was followed by training on administering the questionnaire (data collection guidelines, verification of proper understanding of the questions, role-playing exercises, practice sessions with women, study introduction, and obtaining consent). These training sessions were conducted by the project team. A feasibility study was then conducted over 15 days in February 2024 to assess the feasibility of recruiting participants by telephone (e.g. refusal rate, questionnaire duration, acceptability, comprehension, debriefing, correction of missing values, etc.).

The structured questionnaire consisted of 5 sections:

1. BF practices.

2. Sociodemographic and economic characteristics.

3. Reasons for discontinuing breastfeeding.

4. Prenatal care and health-related behaviors during pregnancy.

5. Obstetrical and newborn characteristics.

Primary outcome

The primary outcome measure for this analysis was the “exclusive breastfeeding at 4 months” (EBF-4) indicator, which corresponds to the proportion of children born in the past 12 months who were breastfed at 4 months of age.

Rationale for the indicator: According to current French recommendations from the National Nutrition and Health Program, it is recommended to introduce solid, semi-solid, and soft foods between 4 and 6 months of age [11–13]. Thus, in accordance with these recommendations, exclusive breastfeeding is advised until at least 4 months of age. While the WHO recommends exclusive breastfeeding until 6 months of age, the 4-month threshold is the national benchmark used in French surveillance systems, including the Epifane survey [20]. This allows for a direct comparison with French regional and national data. Consequently, this threshold is more appropriate for contextualizing these public health results in French Guiana, especially since professional practices in the territory align with those of the PNNS rather than those of the WHO.

Definition of indicators

Table 1 lists all the indicators and their definitions, along with their associated short names, which will be used throughout the rest of the article [10].

Table 1.

Summary and definitions of indicators, ANJE Nutri survey, French Guiana, 2024

A Breastfeeding indicators
Indicator	Short name		Age group	Definition
Ever breastfed	BF-12		Children born in the last 12 months	Percentage of children born in the last 12 months who were ever breastfed
Early initiation of breastfeeding	EIBF		Children born in the last 12 months	Percentage of children born in the last 12 months who were put to the breast within one hour of birth
Breastfeeding during the maternity stay	BF-MS		Children born in the last 12 months	Percentage of children born in the last 12 months who were fed with breast milk from birth through discharge from the hospital
Ever exclusively breastfed	EBF-12		Children born in the last 12 months	Percentage of children born in the last 12 months who were ever exclusively breastfed
Exclusive breastfeeding during the maternity stay	EBF-MS		Children born in the last 12 months	Percentage of children born in the last 12 months who were exclusively breastfed from birth through discharge from the hospital
Exclusive breastfeeding at 4 months of age	EBF-4		Children aged 4 months	Percentage of children aged 4 months who are exclusively breastfed
Exclusive breastfeeding at 6 months of age	EBF-6		Children aged 6 months	Percentage of children aged 6 months who are exclusively breastfed
Continued breastfeeding at 12 months of age	CBF-12		Children aged 12 months	Percentage of children aged 12 months who were fed breast milk during the previous day
Mixed milk feeding under one month	MixMf-1		Children born in the last 12 months	Percentage of infants born in the last 12 months who were fed infant formula, in addition to breast milk, before they were one month old

B Other variables
Terms	Short name	Period		Definition
Maternity stay	-	-		Includes birth and immediate postpartum period in hospital (period up to discharge from the maternity stay)
Maternity supplement	MS	Maternity stay		refers to the occasional consumption, in addition to breast milk, of infant formula, water (sweetened or unsweetened), herbal teas, or any other liquid given during the first few days in the maternity ward. This practice applies to women who are breastfeeding.
Professional breastfeeding support during the maternity stay*	BS-MS	Maternity stay		Percentage of women who self-reported receiving professional breastfeeding support from a healthcare provider in the maternity stay
Professional breastfeeding support after discharge from the maternity stay*	BS-12	Discharge from the maternity stay – 12 months		Percentage of women who self-reported receiving professional breastfeeding support from a healthcare provider after discharge from the maternity stay

*Breastfeeding support (BS-MS and BS-12) provided by healthcare professionals corresponds to the participants reported experiences based on their needs. It refers to the support they felt during their maternity stay and after discharge, provided by the care teams

Statistical analysis

The statistical analysis was performed using the svy functions of STATA 19.5 software, Stata Corporation, Texas, USA.

After verifying the quality of the data collected, a descriptive analysis was performed to summarize the characteristics of the study population. The various IYCF indicators were estimated for qualitative variables in the form of frequencies and percentages, with 95% confidence intervals (CI), and for quantitative variables using a mean and standard deviation or a median and interquartile range (IQR) depending on their distribution.

The data were weighted using the CALMAR macro (SAS software) from the National Institute of Statistics and Economic Studies (INSEE) by calibrating margins based on reference data from the RIGi (n = 7,690) [23]. This calibration, performed using the raking ratio method, considered the mother’s age group (3 categories), place of birth (9 categories), parity (4 categories), hospital delivery center (3 categories) and mode of delivery (3 categories) to ensure the representativeness of the sample. The quality of the weighting was assessed by examining the calibration diagnostics: the efficiency was 78%, the weight ratio remained below the recommended threshold of 4, and convergence was achieved after six iterations. Due to the weighting of the survey, all results are presented as weighted proportions rather than numbers per category.

Bivariate analyses were tested using the Chi² test (or Fisher’s exact test) to explore all crude associations between the independent variables (socio-demographic characteristics of the mother, child or delivery) and EBF-4. These variables were selected based on background knowledge, either from the literature or from our understanding of the context in French Guiana.

Variables with a p-value ≤ 0.2 in the bivariate analysis were considered for inclusion in the multivariate model.

Then, a multivariate analysis was then conducted to control for confounding and identify factors associated with EBF-4. The modeling strategy aimed to create an explanatory model. Candidate variables were included in the model following the bivariate analysis by selecting variables up to a significance threshold of 0.2. No backward or stepwise selection procedure was applied. The decision was then made to retain the full model in order to obtain adjusted effect estimates and minimize residual confounding. To avoid collinearity between variables included in the full multivariate model, variance inflation factors (VIF) were examined: the mean VIF for all included variables was < 2.5. Finally, based on scientific plausibility and previous literature, interactions between maternal age, maternal birthplace, educational level, and nutritional supplementation were explored individually.

The chosen multivariate model was a modified Poisson regression model. When the prevalence of a binary event exceeds 10% (here, the rate of EBF-4), the odds ratio (OR) derived from logistic regression overestimates the strength of the association compared to the relative risk (RR). This overestimation can lead to a misinterpretation of the results. Thus, the prevalence ratio (PR) estimated by the modified Poisson regression is directly interpretable and more relevant [24].

Finally, missing values were handled as follows:

Descriptive, subgroup, and bivariate analyses were performed using complete-case data. Multivariate analyses were conducted both on the original (complete-case) data and after multiple imputation of missing values.

Imputation was performed on three variables in the multivariate model that had missing values (< 10%): obesity, hospitalization, educational level. Missing data were handled using multiple imputation by chained equations (MICE) with the “mi impute chained” command in Stata (version 19.5) [25]. After verifying the assumption of missing at random (MAR), each incomplete variable was imputed using regression models appropriate for its type (logit for obesity (n = 39; 10.2%) and hospitalization (n = 3; 0.8%), ologit for educational level (n = 15; 3.9%), based on the other variables in the dataset. The outcome variable, EBF-4, was included in the imputation model as an auxiliary variable to improve prediction of missing covariates. In total, 20 imputed datasets were generated with a sufficient number of iterations to ensure convergence. The estimates were combined across imputations using Rubin’s rules to account for intra- and inter-imputation variability [26, 27].

The weights derived from the CALMAR calibration were incorporated using the “mi estimate: svy” command.

The multivariate model is thus presented before and after imputation of missing values. The complete-case model includes 330 participants and reflects the exclusion of observations for which any variable in the multivariable model had missing values. Following multiple imputation, the analytical sample size increased to 374. However, this does not represent the full sample size of 382, as one variable (relationship status, n = 8, 2.1%) could not be imputed due to the absence of suitable predictor variables. The eight participants with missing values for this variable were excluded from both models.

Results

Representativeness of the sample

From the stratified and proportional sample of 786 women included in the initial Nutri Pou Ti’moun study [18], 382 mother-child pairs were selected for the ANJE-Nutri study (ancillary study). The participation rate in the study was 48.6%.

The refusal rate was 8.8% (69/786). For 16 of the women included in the primary study, neither a phone number nor socioeconomic data were found in the administrative records. The main reason for exclusion among the 770 participants was failure to answer phone calls in 46.4% of cases. The second most common reason was an invalid phone number. This data has been summarized in the flowchart shown in Fig. 1.

Fig. 1.

Flow chart, ANJE Nutri Survey, French Guiana, 2024. *CHC: Cayenne Hospital Center (French: Centre Hospitalier de Cayenne). **CHOG: Western French Guiana Hospital Center (French: Centre Hospitalier de l’Ouest Guyanais). ***CHK: Kourou Hospital Center (French: Centre Hospitalier de Kourou)

To assess potential non-response bias, socioeconomic comparisons were conducted between women included in the study and those not included (Appendix 1). To further evaluate the representativeness of the sample, additional comparisons were performed between included women and those recorded in the regional birth registry (RIGi) (Appendix 2).

Appendix 1.

Representativeness of the sample of women who participated in Nutri Pou Ti’moun, included in ANJE Nutri (n = 382), compared to women not included (n = 388)*

Characteristics of the ANJE Nutri sample
		Included	Not included**	p
		%	%
Mother’s age (in years)	< 20	5.2	13.7	0.000
	20–34	71.5	67.5
	≥ 35	23.3	18.8
Mother’s place of birth	French Guiana	38.9	46.0	0.000
	Mainland France	6.7	4.2
	Other overseas departments and regions	1.6	1.7
	Suriname	10.2	18.1
	Brazil	5.1	7
	Haiti	33.5	15.3
	Guyana	0.8	1.7
	Dominican Republic	1.6	1.9
	Others	1.6	4.2
Living area	French Guiana	61.7	42.6	0.000
	Western French Guiana	29.8	43.2
	Savannah	7.5	9.5
	Eastern Guiana	1.1	4.7
Parity (before delivery)	Primiparous	30.6	34.1	0.363
	Multipara (1 to 9)	69.4	65.9
Level of education	≤ primary	18.5	11.5	0.000
	Secondary	60.5	75.0
	Higher	21.0	13.5
Hospital delivery center	CHC	63.4	47.4	0.000
	CHK	7.9	9.3
	CHOG	28.8	43.3

* The analysis covers 770 women rather than 786 because for 16 of the women included in the primary study, neither a phone number nor socioeconomic data were found in the administrative records

**Non-inclusions: Refusals, Unreachable, deceased child, absent mother

Appendix 2.

Representativeness of the sample of women who participated in Nutri Pou Ti’moun, included in ANJE Nutri (n = 382), compared to women included in RIGi 2021 (n = 8,152)

ANJE Nutri vs. RIGi
		ANJE Nutri (2024)	RIGi (2021)	p
		%	%
Mother’s age (in years)	< 20	5.2	12.1	0.000
	20	71.5	67.1
	≥ 35	23.3	20.8
Mother’s place of birth	French Guiana	38.9	42.6	0.000
	Mainland France	6.7	3.7
	Other overseas departments and regions	1.6	0.9
	Suriname	10.2	18.2
	Brazil	5.1	6.5
	Haiti	33.5	22.4
	Guyana	0.8	1.3
	Dominican Republic	1.6	2.0
	Others	1.6	2.5
Living area	French Guiana	61.7	44.0	0.000
	Western French Guiana	29.8	44.2
	Savannah	7.5	9.0
	Eastern Guiana	1.1	2.9
Parity (before delivery)	Primiparous	30.6	27.9	0.186
	Multipara (1 to 9)	69.4	72.1
Hospital delivery center	CHC	63.4	49.9	0.000
	CHK	7.9	9.5
	CHOG	28.8	40.7

In both comparisons, participation was lower among women under 20 years of age. The sample also showed an overrepresentation of women of Haitian origin and of those who delivered at the reference maternity hospital in Cayenne. In the comparison with non-participants (Appendix 1), the sample additionally included a slightly higher proportion of women with higher educational attainment.

These differences supported the use of post-stratification weighting.

Prevalence and description of breastfeeding practices

In 2024, 91.6% of children born in the last 12 months were ever breastfed, including 89.8% in the maternity stay (65.8% within one hour of birth). Considering that the duration of BF for children still being breastfed at the time of the questionnaire was 12 months, the median duration of breastfeeding was 6 months (interquartile range IQR: 3.0–12.0; minimum of 1.0 day).

Among children born in the last 12 months, 51.6% were ever exclusively breastfed, including 20.1% in the maternity stay. The majority began EBF after leaving the maternity stay. The median duration of EBF among children born in the last 12 months who were ever breastfed was 3.0 months (interquartile range IQR: 1.5–4.9; maximum of 7.0 months).

Among children who were ever breastfed in the last 12 months, 77.4% had received a maternity supplement (MS), mainly infant formula (75.5%).

In the sample (n = 382), 69.6% of women reported receiving breastfeeding support from a professional at the maternity stay (BS-MS). After leaving the maternity ward, 40.9% of women (n = 176/349) received breastfeeding support (BS-12) from private practitioners (79.6%), a healthcare professional in a hospital setting (14.5%), a maternal and child health center (4.8%) or associations (0.3%). Among those who did not receive BS-12 support, 22.8% wished they had.

Regarding our variable of interest, 22.3% of the children were exclusively breastfed at 4 months of age (EBF-4).

Table 2 describes the infant’s feeding methods.

Table 2.

Infant feeding methods (N = 382), weighted data, ANJE Nutri survey, French Guiana, 2024

Breastfeeding indicators	Prevalence rates
	% [95% CI]
Ever breastfed (BF-12)	91.6 [88.0–94.2]
Early initiation of breastfeeding (EIBF)	65.8 [60.2–71.0]
Breastfeeding during the maternity stay (BF-MS)	89.8 [86.1–92.7]
Ever exclusively breastfed (EBF-12)	51.6 [45.9–57.2]
Exclusive breastfeeding during the maternity stay (EBF-MS)	20.1 [16.6–25.6]
Exclusive breastfeeding at 4 months of age (EBF-4)	22.3 [17.8–27.6]
Exclusive breastfeeding at 6 months of age (EBF-6)	9.7 [6.8–13.6]
Continued breastfeeding at 12 months of age (CBF-12)	29.7 [24.8–35.0]
Mixed milk feeding under one month (MixMf-1)	48.7 [43.1–54.4]

The table in the appendix 3 describes the reasons for stopping BF according to the age of the young child. The “Other” category includes other reasons cited by mothers, such as the child’s teething, family deaths or a new pregnancy. Regardless of the duration of breastfeeding, the most frequently cited reason by mothers was the feeling that the child no longer wanted to breastfeed.

Appendix 3.

Reasons for stopping breastfeeding, according to age at cessation, weighted data, (in %, multiple responses, n = 224), ANJE Nutri survey, French Guiana, 2024

Reasons for stopping breastfeeding (multiple choice response)	n (%)	Distribution by age at cessation
		Stopped ≤ 2 months (n = 94)	2 < cessation ≤ 6 months (n = 112)	cessation > 6 months (n = 177)
My child no longer wanted to	108 (45.5)	60.8	43.2	34.0
Returned back to work	29 (13.4)	6.4	13.6	20.2
Impression of lack of milk	27 (9.7)	14.6	11.5	1.2
Did not wish to continue	21 (11.7)	7.0	14.0	12.1
Difficulty breastfeeding	19 (7.5)	6.5	8.5	6.4
Fatigue	8 (4.1)	6.0	2.1	6.0
Breast disease/pain (mastitis, cracked nipples, etc.)	7 (2.6)	4.7	2.0	1.0
Medical indication (disease, risk of transmission, etc.)	6 (2.1)	2.1	2.5	1.2
Request from my spouse	1 (0.2)	0.0	0.4	0.0
Others	24 (13.3)	4.4	12.8	23.7

According to bivariate analyses (Table 3), EBF-4 was significantly more common among mothers under the age of 20 and varied by hospital delivery (Table 3.A). It occurred significantly less frequently in cases of body mass index (BMI) greater than 30 (Table 3.B) and in cases involving any MS (Table 3.C).

Table 3.

Bivariate analysis: exclusive breastfeeding EBF-4, weighted data, ANJE Nutri survey, French Guiana, 2024

A Socio-demographic and economic characteristics of the mothers
Variables						BF-12*			Median durations of breastfeeding (months)			EBF-4**
												No			Yes
												%			%			p-value
Age*** (years) (n = 382)		< 20				94.6			6			50			50			0.0018
		20–34				91.4			6			83			17
		≥ 35				90.7			8			77			23
Place of birth (n = 382)		French Guiana				92.0			6			83			17			0.0704
		Mainland France and overseas departments				89.8			6			57			43
		Haiti				89.7			11			79			21
		Other				92.8			8			73			27
Parity (n = 382)		Multiparous				90.7			6			80			20			0.1629
		Primiparous				94.2			6			72			28
Hospital delivery center**** (n = 382)		CHC				91.8			6			74			26			0.0524
		CHK				86.6			6			65			35
		CHOG				92.5			6			85			15
Education level (n = 367)		≤ primary				91.4			6			80			20			0.1608
		Secondary				90.3			6			79			21
		Higher				97.3			6			67			33
Relationship status (n = 374)		Single				88.3			5			83			17			0.0905
		In a relationship				93.1			6			74			26
PUMa***** (n = 356)		No				88.7			6			75			25			0.5714
		Yes				92.7			6			78			22
Administrative status (n = 379)		French citizen				91.4			5			79			21			0.4971
		Foreign national with residence permit				94.1			7			80			20
		Foreign national without residence permit				84.5			9			69			31
		No response				100			11.7			83			17
Having a stable and declared income in the household (n = 371)		No				89.3			6			79			21			0.4403
		Yes				94.3			6			75			25
Household food insecurity****** (n = 371)		No				93.4			6			76			24			0.5485
		Yes				87.6			6			80			20
B Prenatal care and health behaviors during pregnancy
Variables				BF-12*				Median durations of breastfeeding (months)			EBF-4**
											No			Yes
											%			%			p-value
At least one hospitalization (n = 379)	No			91.6				6			76			24			0.1927
	Yes			92.2				6			83			17
Mother’s maternity stay ≤ 3 days (n = 382)	No			87.7				5			76			24			0.5894
	Yes			94.2				6			79			21
At least 3 ultrasounds performed (n = 376)	No			89.2				6			76			24			0.6817
	Yes			92.5				6			79			21
Presence of gestational diabetes (n = 382)	No			91.7				6			76			24			0.1406
	Yes			91.2				4			85			15
*******BMI ≥ 30 (n = 382)	No			92.2				6			74			26			0.0219
	Yes			90.6				5			86			14
Smoking (n = 371)	No			92.5				6			77			23			0.2409
	Yes			62.6				2			92			8
Alcohol consumption (n = 382)	No			90.6				6			76			24			0.2539
	Yes			93.8				6			82			18
C Obstetrical and newborn characteristics
Variables					BF-12*		Median durations of breastfeeding (months)			EBF-4**
										No			Yes
										%			%			p-value
Sex of baby (n = 382)			Girl		92.1		6			77			23			0.8561
			Boy		91.2		6			78			22
Type of pregnancy (n = 382)			Single		91.4		6			77			23			0.2243
			Multiple		100		4			100			0
Mode of delivery (n = 382)			Caesarean section		82.0		4			83			17			0.3202
			Vaginal birth		93.9		6			76			24
Episiotomy (n = 354)			No		93.9		6			77			23			0.7757
			Yes		89.3		3			74			26
Anesthesia (n = 382)			No		94.6		7			77			23			0.7612
			Yes		89.7		5			78			22
Child weight < 2500 g (n = 382)			No		92.2		6			78			22			0.8287
			Yes		84.4		4			80			20
Delivery < 37 weeks gestation (n = 382)			No		92.9		6			77			23			0.6467
			Yes		65.5		4			83			17
Transfer of mother (n = 381)			No		91.5		6			78			22			0.4592
			Continuous monitoring unit		100		> 12			100			0
Transfer of newborn (n = 382)			No		92.8		6			78			22			0.9364
			Yes		82.7		6			78			22
Apgar at 5 min < 7 (n = 382)			No		92.3		6			78			22			0.5586
			Yes		58.7		6			68			32
Maternity supplement (MS) (n = 382)			No		100		11			45			55			0.0000
			Yes		100		5			86			14
Professional breastfeeding support during the maternity stay (BS-MS) (n = 382)			No		66.1		3			81			19			0.4026
			Yes		100		6			76			24

*BF-12: Percentage of children born in the last 12 months who were ever breastfed

**EBF-4: Percentage of children aged 4 months who are exclusively breastfed

***Age: Mother’s age at the time of delivery

**** CHC: Cayenne Hospital Center (French: Centre Hospitalier de Cayenne) CHOG: Western French Guiana Hospital Center (French: Centre Hospitalier de l’Ouest Guyanais) CHK: Kourou Hospital Center (French: Centre Hospitalier de Kourou)

*****PUMa: Universal Health Coverage

******Household food insecurity (FI) was measured using the USDA Food Security Survey Module [28]

*******BMI: Body mass index

According to the multivariate analyses (Table 4), practicing EBF-4 was positively associated with being a mother under the age of 20 and having a higher education. Conversely, this practice was negatively associated with being born in French Guiana and with the presence of MS. Models with and without imputation of missing data produced similar results, with stable estimators.

Table 4.

Multivariate analysis – poisson regression model without interactions: EBF-4 (n = 330), weighted data, ANJE Nutri survey, French Guiana, 2024

		Model without imputation n = 330		Model with imputation n = 374
Variables		EBF-4*		EBF-4
		PR** [95% CI]	p	PR [95% CI]	p
Age*** (years)	< 20	4.2 [2.3–7.7]	0.000	3.5 [1.8-7.0]	0.000
	20–34	1 (ref)		1 (ref)
	≥ 35	1.1 [0.6–1.7]	0.825	1.1 [0.7–1.8]	0.593
Place of birth	French Guiana	0.4 [0.2–0.8]	0.004	0.5 [0.3–0.9]	0.027
	Mainland France and overseas departments	0.9 [0.5–1.7]	0.786	0.9 [0.5–1.7]	0.772
	Haiti	1 (ref)		1 (ref)
	Others	1.1 [0.6–1.9]	0.815	1.0 [0.6–1.8]	0.989
Parity	Multiparous	1 (ref)		1 (ref)
	Primiparous	0.8 [0.5–1.4]	0.455	0.8 [0.5–1.4]	0.414
Hospital delivery center****	CHC	1 (ref)		1 (ref)
	CHK	0.9 [0.5–1.6]	0.738	0.8 [0.5–1.4]	0.509
	CHOG	1.0 [0.6–1.6]	0.873	0.8 [0.4–1.4]	0.382
Education level	≤ primary	1.0 [0.6–1.6]	0.853	1.0 [0.6–1.6]	0.850
	Secondary	1 (ref)		1 (ref)
	Higher	2.0 [1.2–3.4]	0.011	1.9 [1.1–3.1]	0.017
Relationship status	Single	1 (ref)		1 (ref)
	In a relationship	1.0 [0.6–1.6]	0.900	1.1 [0.6–1.8]	0.771
Presence of gestational diabetes	No	1 (ref)		1 (ref)
	Yes	0.9 [0.5–1.6]	0.632	0.8 [0.4–1.5]	0.440
*****BMI ≥ 30	No	1 (ref)		1 (ref)
	Yes	0.7 [0.4–1.2]	0.231	0.7 [0.4–1.3]	0.313
At least one hospitalization	No	1 (ref)		1 (ref)
	Yes	0.9 [0.5–1.5]	0.560	0.9 [0.6–1.5]	0.771
Maternity supplement	No	1 (ref)		1 (ref)
	Yes	0.3 [0.2–0.5]	0.000	0.3 [0.2–0.5]	0.000

* EBF-4: Percentage of children aged 4 months who are exclusively breastfed

** PR: Prevalence ratio

*** Age: Mother’s age at the time of delivery

**** CHC: Cayenne Hospital Center (French: Centre Hospitalier de Cayenne) CHOG: Western French Guiana Hospital Center (French: Centre Hospitalier de l’Ouest Guyanais)

CHK: Kourou Hospital Center (French: Centre Hospitalier de Kourou)

***** BMI: Body mass index

A significant interaction was identified between age and place of birth. However, due to the small number of observations in certain categories, a reliable interpretation remains limited.

Discussion

According to the ANJE Nutri study, 91.6% of children born in the last 12 months were ever breastfed with a median breastfeeding duration of six months. Among these children, 51.6% were ever exclusively breastfed (including 20.1% at birth during their maternity stay), with a median exclusive breastfeeding duration of three months. Among breastfed children, 77.4% received maternity supplement. In the maternity stay, 69.6% of the women included in the study reported receiving breastfeeding support (BS-MS). Upon leaving the maternity stay, 40.9% of breastfeeding mothers reported receiving breastfeeding support (BS-12), while 22.8% would have liked to receive BS-12 but did not. The main reasons for stopping BF were the child’s refusal to breastfeed, the feeling of not having enough milk, the mother’s desire to stop, and a return to work.

EBF-4 was achieved by 22.3% of infants. Maternal factors associated with EBF-4 were young age (under 20), a high level of education (university degree) and not using MS. Mothers born in French Guiana were less likely to practice EBF-4 than those born in Haiti.

Comparison with Mainland France and the overseas departments and regions

As in other overseas departments and regions (DROMs), a higher proportion of BF-12 in French Guiana than in other French regions: 86.2% in Guadeloupe (2021), 91.6% in French Guiana (2024), 92.8% in Martinique (2021) and 93.8% in Mayotte (2019), compared to 77.0% in Mainland France (2021). Similarly, the duration of BF was longer in French Guiana (median of 6.0 months) than in Mainland France (4.5 months in 2021) [20, 29]. Given the level of food and nutritional insecurity among pregnant women in French Guiana, BF practices represents an effective strategy for limiting expenditure on infant formula [18]. Furthermore, in a context of precariousness, limited access to drinking water, particularly for preparing baby bottles, high prevalence of overweight children aged 5 to 14 (17.8%), obesity (14.1% in 2023) and increased exposure to infectious diseases in French Guiana, BF would provide crucial protection against diarrhea and childhood obesity and promote immune development in infants [30–32].

The prevalence of EBF-12 in French Guiana (51.6%) is lower than in Mainland France (59.0%) and represents the lowest rate among the DROMs i.e. between 56.7% and 61.0% in Guadeloupe, Réunion and Saint Martin, 76.6% in Martinique and 80.5% in Mayotte [20, 29]. On the other hand, the median duration of exclusive breastfeeding is twice as long in French Guiana as in Mainland France [20].

Women who gave birth in French Guiana were more likely to report receiving BS-12 (40.9%) than in Mainland France (30.2%) or Martinique (28.9%), but a higher proportion of people in French Guiana than in Mainland France reported that they did not receive BS-12 when they wanted it (22.8% vs. 16.8%) [20].

The prevalence of both breastfeeding and exclusive breastfeeding up to 12 months in French Guiana is encouraging. However, the more frequent practice of mixed milk feeding in French Guiana (48.7%), than in Mainland France (18.0%) or Martinique (17.0%) and the early and frequent use of MS highlight specific challenges and room for improvement in the territory.

Socio-demographic and economic profiles of mothers

The socioeconomic characteristics of the mothers in the study that were significantly associated with EBF-4 were young age, being born outside French Guiana and having completed higher education.

Unlike most studies [33], our research showed that mothers under the age of 20 were significantly more likely to EBF-4. Social support, greater self-confidence, or educational level may have positively influenced breastfeeding outcomes in this age group, as previously observed in certain studies [34, 35]. While it is true that in French Guiana, pregnancies among women under 20 are common (12.5%) [36], some young mothers could nevertheless benefit from stronger intergenerational and family support in caring for their infants, as well as from medical and social support programs specifically designed for teenage mothers [34, 37, 38]. However, this association should be interpreted cautiously because of the small number of mothers in this age group.

According to our findings, the mother’s place of birth was also associated with the practice of EBF-4 in French Guiana. Differences in practices based on place of birth may reflect different traditions and customs. However, due to the existence of different communities among women born in French Guiana (Haiti, Brazil, Aluku, Palikur, Wayanas, Hmongs, Creole, etc.), this result may mask more subtle differences between them.

Finally, the association between high educational attainment and EBF-4 may reflect better health literacy among these women [39], thus facilitating understanding of the benefits and knowledge surrounding exclusive breastfeeding.

Contrary to the literature [40] and what was expected in the context of French Guiana, variables referring to financial insecurity (at least one stable and declared income in the household) and household food insecurity were not significantly associated with EBF-4. Household economic status appears to be less associated with exclusive breastfeeding practices than other variables, a finding that may be explained by the existence of a medical-social safety net in France for pregnant and breastfeed women (social aid, medical aid of the state, free post-natal center).

Reasons for using mixed breastfeeding and support from the maternity stay onwards

According to this study, the use of MS was negatively associated with EBF-4. Moreover, as shown in Table 3.C, the median breastfeeding duration was shorter among infants who received MS compared to those who did not (5 months vs. 11 months, respectively). This associations were demonstrated in particular in a 2019 study conducted in eight French maternity wards, which examined the determinants of breastfeeding difficulties at six months [41]. There may be multiple reasons for using a MS. Indeed, this practice may reflect fears of insufficient weight gain in the newborn, the mother’s doubts about her ability to breastfeed, and a negative social perception of BF (which is restrictive and part of the private sphere) [7]. The use of MS may also reflect breastfeeding difficulties, neonatal illness, a maternal preference for mixed feeding, limited knowledge, ward practices, or clinical indications.

The use of MS, also appears in the literature as a factor associated with the absence of prolonged exclusive breastfeeding. Studies, such as those conducted in Brazil [42], have shown that the use of MS, especially infant formula, was associated with an increased risk of early weaning [43, 44].

Indeed, it has been recognized that using baby bottles, particularly starting in the maternity stay with MS and pacifiers can strongly influence the duration of EBF practices, for both the infant and the mother. Introducing MS can alter an infant’s oral dynamics and interfere with breast suction technique. This may result in latching difficulties and decreased feeding frequency [42, 45].

Additionally, the early introduction of liquids other than breast milk could substantially reduce its protective effects against infectious diseases [1].

However, the possibility of reverse causation should be considered, as the direction of the association remains uncertain. The use of MS may, in many instances, reflect limited support, insufficient breastfeeding knowledge unless it represents an informed decision by the mother [42, 46].

International literature therefore suggests that breastfeeding support from professionals or peers immediately after childbirth has a positive impact on continued breastfeeding [43, 47, 48]. The NutriNet-Santé cohort (2022) suggested that the increase in BF in Mainland France is due to growing awareness of the benefits of BF and improved prenatal and postnatal support policies [49].

The proportion of mothers receiving BS-12 in French Guiana is higher than in Mainland France and Martinique. However, the study also shows that a significant number of women did not receive BS-12 even though they needed it. Several factors may account for this finding. First, women’s experiences and expectations regarding this support may have been influenced by the differing principles of the IYCF across maternity wards in French Guiana and by varying cultural norms. In addition, other factors may come into play, such as unequal access to postnatal care facilities (geographical, socioeconomic, psychological, and language barriers), a lack of coordination between hospitals and local health centers, a lack of health literacy, or significant social isolation. These results, supported by recent meta-analyses, suggest the need for structured support, tailored to women’s socioeconomic profiles and cultural habits, starting during pregnancy and reinforced in the postpartum period (home visits, lactation consultants). This support would be a crucial lever for limiting early cessation and promoting free and informed choices [50, 51]. Just as there may be an inverse causal relationship between MS and BF practices, the observed associations with healthcare professionals support and guidance may also be subject to indication bias. Infants receiving additional support may represent more complex clinical situations or breastfeeding challenges, which could partly explain the observed relationships.

These initiatives address the high reliance on MS. Multiple factors may explain these practices, including hospital policies (lack of certification under the Baby-Friendly Hospital Initiative (BFHI), lack of standardized protocols, infrequent professional training, staff turnover, professional burnout), clinical practices (separation of mother and child, delay in skin-to-skin contact, lack of dedicated time, reaction to the newborn’s weight loss), and individual (social pressure, low sense of self-efficacy, community norms).

To address these challenges, a charter of commitment regarding infant feeding for healthcare professionals across the region has been developed in collaboration with all stakeholders, in accordance with BFHI recommendations, as recently recommended in the PNNS 2026 (page 44–46) [3, 52–54]. The adoption of this charter would be a key factor in harmonizing practices, limiting the non-medical use of maternity supplements, and sustainably improving exclusive breastfeeding rates at the regional level.

Although the causal nature of these relationships has yet to be determined, these findings have guided the implementation of initiatives aimed at promoting best practices in infants feeding in French Guiana. Structured support programs are currently being tested to overcome the identified barriers. These programs include prenatal and postnatal workshops, as well as early individualized support in maternity wards based on health mediation. Improved practices can also be fostered through individualized or group-based approaches grounded in theories such as Bandura’s self-efficacy theory [55, 56].

Strengths and biases of the study

While there was very little data available for this geographical area, this assessment made it possible to update IYCF indicators and compare them with those in other French regions. These results also provided a better understanding of practices in French Guiana and enabled hypotheses to be made about the obstacles and levers for improving BS-MS, BS-12 pathways [20, 29].

Despite potential inclusion biases related to the inability to reach certain mothers without telephones, or due to frequent travel between two countries at the borders with Suriname and Brazil, the survey covered the three maternity wards in the territory and was conducted by multilingual interviewers, thus limiting linguistic bias. In addition, the representativeness of the sample was improved using post-sampling weighting and the robustness and stability of the statistical analyses were demonstrated using several multivariate models (sensitivity analysis).

Data on BF practices were collected via telephone interviews approximately 12 months after delivery. This retrospective collection is likely to introduce a memory bias, particularly regarding the duration of EBF and the timing of the introduction of complementary foods (solid, semi-solid). This bias may lead to an overestimation of EBF duration and misclassification. According to the literature, the margin of error for estimating BF duration in retrospective recalls at 12 months is 2 weeks [57].

Although post-stratification weighting improved the representativeness of the sample with respect to available population characteristics, residual selection bias cannot be ruled out. Adjustment was limited to variables available in the regional reference database (age, place of birth, parity, delivery hospital and mode of delivery), and it was not possible to account for unmeasured factors (e.g., behavioral or contextual characteristics such as motivation to breastfeed, mobility, access to a telephone, or health literacy) or variables not recorded in the database, such as level of education. In addition, a social desirability bias related to self-reported breastfeeding data remains a possibility [58].

Conclusion

A description of breastfeeding practices in French Guiana revealed significant disparities in the support provided to local populations, with early support needs varying according to the maternal profiles identified and highlighting the limitations of standardized approaches in the face of diverse situations. There is a gap between mother’s expressed needs and the available support, which underscores the need for improved coordination across the perinatal care pathway. This coordination should be strengthened by fostering connections between the city and the hospital.

These findings underscore the need for concrete actions to promote informed choices about exclusive breastfeeding in French Guiana. The use of infant formula supplementation in maternity wards should be subject to a service protocol to minimize its use without a medical indication during the early postnatal period. Professional training and evaluations of professional practices should be used to strengthen breastfeeding counseling before discharge. Finally, post-discharge support must be culturally appropriate, linguistically adapted, and tailored to the diverse population of French Guiana.

It should be noted that the rates of BF-12 observed in French Guiana remain encouraging, reflecting a well-established practice and providing fertile ground for perinatal health policies or actions.

The ANJE Nutri study has thus generated relevant results for the analysis of local dynamics in perinatal health. The qualitative component of this study (currently being analyzed) will refine our understanding of breastfeeding practices and support the development of more tailored interventions for mothers and infants.

Abbreviations

BF

Breastfeeding

BFHI

Baby-Friendly Hospital Initiative

BMI

Body mass index

CHC

Cayenne Hospital Center (French: Centre Hospitalier de Cayenne)

CHK

Kourou Hospital Center (French: Centre Hospitalier de Kourou)

CHOG

Western French Guiana Hospital Center (French: Centre Hospitalier de l’Ouest Guyanais)

CI

Confidence interval

DROMs

Overseas departments and regions (French: Départements et Régions d’Outre-mer)

EBF

Exclusive breastfeeding

IBCLC

International Board-Certified Lactation Consultant

INSEE

National Institute of Statistics and Economic Studies (French: Institut national de la statistique et des études économiques)

IYCF

Infant and young child feeding

MCP

Maternal and child protection

MICE

Multiple imputation by chained equations

OR

Odds ratio

PNNS

National Nutrition and Health Program (French: Programme National Nutrition Santé)

PR

Prevalence ratio

RIGi

Computerized pregnancy outcome register (French: Registre d’Issue de Grossesse informatisé)

UNICEF

United Nations International Children’s Emergency Fund

VIF

Variance inflation factors

WHO

World Health Organization

Author contributions

CB, MG, NH and NC conceptualized the study, LM, SM, EL, and YA contributed to the survey and data collection, LM and CB analyzed the data, LM, CB, CG, MG, CD, and MN contributed to interpretation, LL and RA provided clinical expertise in maternal health and breastfeeding and LM drafted the manuscript. All authors approved the final version of the manuscript.

Funding

This study received public funding from the French Guiana Regional Health Agency.

Data availability

The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethical approval

The ANJE Nutri study was approved by the South-East 1 Human Research Ethics Committee on February 14 (ID-RCB: 2024-A00035-42 (ID-RCB Principal project 2022-A01772-41); NCT0563128 (NCT Nutri Pou Ti Moun) and complies with CNIL reference methodology 003 (MR003). This study complies with the ethical principles of the World Medical Association–s Declaration of Helsinki — Ethical Principles for Medical Research Involving Human Subjects.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.