1. Introduction
Pregnancy and the postpartum period constitute an important stage in women’s health, in which a series of events take place that require special attention and monitoring by the health system. Although it is a physiological process, it involves a continuum of decision-making in which women need to have sufficient information so that these decisions protect and promote not only their health, but also that of their children.
One of the most important decisions to be made is regarding the feeding the infant will receive. International organisations such as the World Health Organisation and UNICEF recommend exclusive breastfeeding (EBF) for the first six months of an infant’s life and breastfeeding with complementary foods until at least two years of age [
1]. The promotion of EBF is an international target in different programmes such as the Comprehensive Implementation Plan on Maternal, Infant, and Young Child Nutrition of the World Health Assembly [
2], the United Nations Decade of Action on Nutrition 2016–2025 [
3], and the investment framework of the World Bank [
4]. However, despite multiple efforts to protect breastfeeding (BF), rates of EBF at six months of infant life remain very low, at around 38% globally [
5]. Furthermore, laws to protect breastfeeding remain inadequate in most countries [
6]. In Europe, the six-month EBF rate is around 25% [
7]. However, in Spain, the six-month EBF rate has varied from 16.8% in 2019 [
8] to 39% in 2017 [
9]. The data need to be interpreted with caution as the variation in these data is caused by the absence of a unified approach for collecting and monitoring BF information in Spain.
The premature discontinuation of breastfeeding is a complex phenomenon that is influenced by a multitude of factors, including demographic characteristics (e.g., young maternal age, low levels of education and socio-economic status), social considerations (e.g., inadequate workplace support), psychological determinants (e.g., maternal intentions before birth, self-assurance, and engagement in breastfeeding), as well as biological considerations (e.g., infant health concerns, maternal health issues, first-time motherhood, and issues related to lactation) [
10,
11,
12,
13]. These considerations contribute to the multifaceted nature of early breastfeeding cessation. However, several studies have shown that, in many cases, early weaning occurs due to maternal decisions or perceptions, which do not always correspond to reality [
14]. In the face of these false perceptions, health literacy (HL) has a fundamental role because the primary outcome of having a good level of HL is the ability to make good decisions that promote and protect health [
15].
Various authors have broadly defined the concept of HL over time [
16]. Despite the lack of consensus on constructing a single definition of this concept, most authors agree that it is multidimensional, complex, and heterogeneous [
17]. Sørensen et al. proposed an integrated model of HL that looked at cognitive and social skills that enable the individual to address four competencies (access, understand, appraise, and apply health information) and three domains in which the individual interacts with the health system (health care, disease prevention, and health promotion) [
18].
This complex concept of HL has been reformulated and adapted to specific health areas or populations. As a result, it is possible to retrieve a multitude of validated instruments that allow us to generically assess the level of HL, such as the European Health Literacy Survey Questionnaire (HLS-EU-Q) [
19] or the test of functional health literacy in adults (TOFHLA) [
20]. There are also instruments available that focus on measuring literacy in specific health areas, such as the Literacy Assessment for Diabetes (LAD), which addresses diabetes literacy [
21]. Others focus on specific populations, such as the eHealth Literacy Scale (eHEALS), which addresses electronic health literacy in a young population [
22]. It is also possible to retrieve the Maternal Health Literacy Inventory in Pregnancy (MHELIP) instrument, which is designed to measure maternal health literacy [
23]. However, to our knowledge, no previous instrument has measured breastfeeding literacy (BFL).
Recent studies have suggested that an adequate level of HL may be a protective factor against early BF cessation [
12,
13,
24]. However, these studies use generic HL instruments to determine the relationship between HL and the specific health domain of BF. Specifically, they use the Short Assessment of Health Literacy for Spanish-speaking Adults (SAHLSA) [
24] and the Newest Vital Sign (NVS) in its validated version for Spanish-speaking populations [
12,
13,
24]. The main findings of using a generic instrument to explore a particular area of health lack specificity and concreteness in the results obtained, so the authors agree on the need for a specifically validated instrument to measure the level of BFL in women during the perinatal stage [
12,
13,
24].
Therefore, this study aims to design and validate a specific instrument to measure the level of BFL.
4. Discussion
The BLAI presents adequate psychometric properties to assess BFL levels in women during the perinatal period, with adequate construct validity and internal consistency. The exploratory factor analysis explains 60.54% of the variance with four domains, coinciding with the four dimensions covered by the concept of HL (Access, Understand, Appraise, and Apply) developed by Sørensen et al. [
18].
It is worth mentioning that, during the instrument’s modelling, a number of items had a slightly higher loading in dimensions for which they were not designed. However, after thoroughly examining each item to evaluate the feasibility of assigning it to alternative dimensions, the research team determined that it was more appropriate to retain these items within their original dimensions, as the theoretical alignment was more convincing in these dimensions. In addition, two items were removed (Access6, Appraisse6) due to their poor factor loadings. The internal consistency of the BLAI slightly increased after their deletion.
As for the dimensionality study of the instrument, the EFA was run without determining a number of factors to extract, allowing the statistical programme to determine the number of factors based on the Kaiser criterion of eigenvalues greater than 1 [
29]. This is the default method in the statistical programme used, and it is possible to retrieve scientific evidence that casts doubt on its practical usefulness, as has been reported by other authors [
32,
33]. However, the resulting factor structure coincided with the number of dimensions for which the instrument was created. Today, there are other, more commonly used methods to corroborate the appropriate number of factors, such as parallel analysis or the ratio of the first-to-second eigenvalue. However, we have not found a universally accepted criterion. For example, in the case of eigenvalues, there is no criterion for the ratio to be accepted, some authors propose four [
34], others five [
35], but none seem to be based on empirical reasoning. Therefore, it is essential that future studies consider other analyses for studying dimensionality.
While it is true that the use of a single criterion may lead to an overestimation or an underestimation of the actual number of factors, over-extraction leads to fewer measurement errors [
36]. Moreover, it would not be appropriate to treat as unidimensional a construct of which the theoretical foundation is based on more than one factor, even if the multidimensionality is moderate. In the present instrument, an overall score of the construct would tend to lean towards the mean of the possible score range, and would not allow for discerning which competence/s the subject presents, and which others lower the mean score of the construct and would need to be addressed by a practitioner. Therefore, treating the construct in an unidimensional way would diminish its usefulness in practice. However, in order to obtain an instrument with a solid factor structure supported by theoretical and statistical reasoning, it is of utmost importance to progress with the validation process, with larger samples and different methods of studying dimensionality, in order to confirm or refute the factor structure that supports the theoretical reasoning.
In terms of the percentage of variance explained by each of the factors, it can be seen that the Access dimension is the one that explains the highest percentage of variance, followed by the Apply dimension. This may be because these dimensions are more manageable for women, while the Understand and Apply dimensions may be more complex due to the reflection involved in these situations. In other words, the general population can access information related to a given topic (Access) and apply the information they have accessed (Apply). However, people who are not experts in an area may find it more challenging to reflect on whether they adequately understand the information they have accessed (Understand), as well as to assess whether the source of information is reliable or may contain information that is not scientifically supported (Appraise). It is important that this finding is taken into account when addressing any health education, specifically in the area of BF, with the aim of training mothers-to-be, and even health professionals, to reflect on the information accessed in order to increase confidence when making health decisions based on the knowledge they have acquired. Future studies could address this necessary line of research.
As evidenced by the findings of this study, the BLAI questionnaire demonstrates utility in identifying areas where perinatal women may require additional competencies to access, understand, appraise and apply information about BF, not only for self-care purposes but also to prevent occurrences that may impede BF, as well as to foster successful initiation and continuation of BF. Similarly, it would be interesting in future studies to use the BLAI questionnaire to measure the effect of BF training or antenatal education on BF. Similarly, future studies should consider confirmatory factor analysis to confirm the current four-dimensional factor structure, as the evidence does not recommend using the same sample to address all validation phases of a newly created instrument, as this would lead to optimistic results [
37]. In fact, we are currently continuing to collect data in order to be able to carry out the confirmatory factor analysis. However, this is the first publication derived from the design and validation of the instrument based on solid theoretical reasoning, so it is interesting to make its existence known, as well as its first psychometric properties.
This study is a continuation of previous studies that addressed the relationship between health literacy measured by generic instruments and BF [
12,
13,
38]. It has not been possible to retrieve in the literature another validated instrument to address the level of BFL, which makes it challenging to contrast results in the present study. On the one hand, concerning age, the present study did not find a statistically significant association with the level of BFL, in line with the results of Vila-Candel et al., in which the study also showed no significant association with the level of LH [
12]. On the other hand, Valero-Chillerón et al. did find that the mean age among mothers with an adequate level of BFL was higher than those with a limited level of BFL [
13].
In the present study, no statistically significant association was observed between educational level and BFL level in any of the dimensions that comprise the questionnaire, in contrast to other studies that obtained such an association between HL level measured with generic instruments and educational level [
12,
13]. This may be due to the fact that two completely different phenomena; the level of education academically trains you in a certain area, whereas the level of breastfeeding literacy explores the individual’s ability to access information related to breastfeeding, understand that information, evaluate the quality of the information accessed, and apply that information in the specific area of breastfeeding. It is possible that a higher level of education may enhance an individual’s competence in certain areas of daily life, but it may not be sufficient to establish statistically significant relationships across all dimensions of the BFL concept. Another discrepancy is observed for parity. In the present study, a significant association was observed between Appraise BFL and the number of children; whereas this association was not significant in previous studies for HL levels [
12,
13]. In addition, Valero-Chillerón et al. observed an association between the country of origin and the level of HL, while this association could not be observed in the present study regarding the level of BFL, perhaps due to the low participation of women whose country of origin was not Spain [
13]. In line with the findings of Sørensen et al., low socioeconomic status is related to low levels of HL, and, as in the present study, with Inadequate Access BFL [
38].
It has not been possible to retrieve any study in which a statistically significant association was found between HL level measured with a generic instrument and maintenance of EBF at six months. However, Vila-Candel et al. did find a statistical association between LH level and maintenance of EBF at one, two, and four months of infant life, although they did not re-measure at six months [
12]. Moreover, all studies seem to confirm the multi-causality derived from early breastfeeding cessation [
12,
13,
24,
39]. This is why it may not be appropriate to address this relationship using a generic instrument to give sufficient weight to the level of HL on the duration of EBF, and it may be advisable to use a specific instrument to assess the level of BFL. Future studies should address this aspect to confirm the results obtained.
It was observed that the percentage of women who opted for mixed breastfeeding had the lowest percentage of adequate understanding and adequate Apply BFL. This may be a chance finding due to the limited percentage of this category in the present study. Contrasting these results in future studies conducted with larger samples would be interesting. It is worth mentioning that the rates of EBF and mixed feeding are similar to those reported in the study by Chertok et al., and point to an increase in the numbers of mixed breastfeeding and formula feeding after the SARS-CoV-2 pandemic, due to the lack of support for breastfeeding during the pandemic, among other factors [
40].
We must recognise several limitations in our study and cautiously interpret the results. Firstly, it should be noted that since we could not retrieve any previous instruments that measure the level of BFL or any other measurement method that could be used as a gold standard reference, it was not possible to study convergent validity. Secondly, it was challenging to randomise the study sample, so convenience sampling was used. Thirdly, it is necessary to advance the process of analysing the dimensionality of the instrument. The methods used in the present study need to be tested against more objective criteria in larger samples, minimising additional survey items to the BLAI questionnaire to try to avoid possible response bias among participants, in order to confirm the factor structure.
Despite the limitations, we believe that the good psychometric properties of the instrument suggest that its use should be considered, as it is the first validated instrument to measure the level of BFL. Previous studies have found that the percentage of women with limited HL was significantly higher among mothers who did not reach four months [
12] or six months of EBF than among those who did reach EBF at these follow-up points [
13,
24]. Therefore, it is interesting to study the relationship between the level of BFL using the BLAI questionnaire and maintenance of EBF at six months, as well as to study the explanatory power of the instrument. Future studies will also allow us to contrast the results obtained and explore the possibility of refining the instrument or the suitability of maintaining the current version. Similarly, future studies could adapt and validate the current version of the instrument among health science professionals and students.