**1. Introduction**

Currently, there is extensive evidence of the teratogenic effects of prenatal alcohol exposure, which can translate into a broad spectrum of abnormalities that make up what is known as fetal alcohol spectrum disorder (FASD) [1,2]. The damage caused to the forming nervous system is permanent and has a number of consequences for the biological development of the fetus as a whole, as well as for subsequent neurocognitive and social development in childhood and the other stages in lifespan [3]. FASD therefore represents a major health and socio-educational problem, since it is

the main non-hereditary, avoidable cause of learning difficulties in developed countries, with serious consequences for new-born babies that will last their entire lives [4–6].

It is uncertain how prevalent alcohol consumption during pregnancy is, due to the tendency for it to be underestimated when it is assessed using scales, and as a result of the lack of studies with biomarkers that allow it to be estimated reliably. However, there are worrying data available that tell us that alcohol consumption is a very widespread practice among expectant mothers (particularly in Europe, North America, Australia, and in countries such as South Africa). Its incidence worldwide has been estimated at 9.8% [7]. In a study carried out on expectant and new mothers found online from eleven European countries, 15.8% of them stated that they consumed alcohol during pregnancy, with the United Kingdom (28.5%) and Russia (26.5%) being the countries with the highest estimated level of prevalence in this study [8]. In parallel to consumption, the countries that have the highest estimated prevalence of FASD (19.8 per 1000 population) are those that belong to the European region of the WHO [9]. All of this makes FASD a global public health problem, requiring effective strategies for prevention and early diagnosis, and representing a crucial challenge for healthcare personnel in general, and obstetricians and midwives in particular.

Health professionals play a very important role in providing preventive advice regarding healthy lifestyles in the periconceptional period, during pregnancy, and postnatally. However, there are signs that large sections of healthcare professionals (general practitioners (GPs), obstetricians, and midwives) are not doing their job fully and properly in this regard. The health advice that pregnant women receive about the risks inherent in consuming alcohol during pregnancy frequently proves contradictory, or else it does not reach those with a lower level of education effectively [10]. A number of studies suggest that not having received any specific training in this area could explain why not all professionals routinely inquire about alcohol consumption when caring for pregnant women, or do not always provide appropriate information on the subject [11,12].

Another factor which might explain this is the ambivalence of the official guidelines themselves in this area, or their inconsistency from one country to another, and over time. A review of the Australian and American guidelines related to alcohol consumption during pregnancy, carried out by Whitehall in 2006 [13], highlighted the fact that there was a certain permissiveness shown by the health authorities regarding alcohol consumption during pregnancy. They even failed to recommend abstention, arguing that it might cause disproportionate anxiety and therefore prove even more harmful than alcohol consumption. At the same time, another review of the policies and guidelines on alcohol consumption during pregnancy in English-speaking countries [14] showed that these varied from country to country and within the countries themselves. With the passing of time, since it has not been possible to establish that there are safe levels of alcohol consumption during pregnancy, and at the same time, since there is increasing evidence that low levels of alcohol consumption can lead to risks to fetal development [15], there are many countries where official health guidelines on pregnancy recommend abstaining completely from drinking alcohol during this time. Thus, since 2002, France has recommended total abstention from alcoholic beverages during pregnancy [16]. National public health bodies in Australia [17], Denmark [18], and Norway [19] have made equivalent recommendations. At the same time, in Scotland, since 2012, the Chief Medical Officer has advised that "pregnant women and those trying to conceive should avoid alcohol" [20]. Equally, in the USA, the Message to Women from the U.S. Surgeon General stated "No amount of alcohol consumption can be considered safe during pregnancy" [21]. At the same time, the American Academy of Pediatrics recommends that health professionals promote total avoidance of alcohol consumption throughout pregnancy, in line with the principle of precaution [4]. In Canada, the 2010 consensus guidelines from the Society of Obstetricians and Gynaecologists of Canada support alcohol abstention during pregnancy [22].

In order to be able to develop effective FASD prevention, the starting point must be a good understanding of the current situation of the problem. It is particularly crucial to know about the factors that cause or encourage expectant mothers to consume alcohol during pregnancy.

There are some studies which aim to identify the factors that predict alcohol consumption during pregnancy. Several of them have concluded that the most important factor is the frequent consumption of alcohol prior to pregnancy [23–27]. Bearing this in mind, it is particularly worrying that in countries such as Spain, at present, two thirds of women of child-bearing age consume alcohol. According to the Spanish National Health Survey of 2017 [28], only 36.96% of women aged 15–24 years of age and 23.78% aged between 25 and 34 identified themselves as non-drinkers. These data present us with a potential problem in the years to come—an increased incidence of FASD—unless effective healthcare provided during the periconceptional period helps them to stop consuming alcohol.

Other predictive factors identified in some studies are having been the target of violence [25], high socioeconomic status, an unplanned pregnancy, and late childbearing age [29]. Similarly, smoking or using other drugs prior to becoming pregnant prove to be predictors of alcohol consumption during pregnancy [26,30,31].

Research to date into the factors that predict alcohol consumption during pregnancy is limited and some yield contradictory results. There are also some potential predictors that have scarcely been explored, such as obstetric history, the partner's alcohol consumption, health advice received regarding alcohol consumption during pregnancy, and the perception of damage resulting from prenatal exposure to alcohol. These factors, in real life, presumably do not act in isolation, but rather interact with one another. To our knowledge, the interaction between a wide range of predictive factors of alcohol consumption during pregnancy has not been studied to date. It is also particularly important to identify these predictors and the interactions between them in those regions of the world where there is a combination of a high rate of alcohol consumption among women of childbearing age and limited implementation of healthcare programs aimed at FASD prevention. The case of Spain, just like that of other European countries, may be particularly illustrative for all regions of the world where these circumstances exist.

Therefore, our study, conducted using a sample of expectant mothers who attended a routine pregnancy check-up in a city in the south of Europe, aimed to determine a wide range of factors that predict alcohol consumption during pregnancy and to identify the relative weight of each of them. Additionally, the study was also designed to assess the degree of interaction between different factors (sociodemographic factors, obstetric history, the partner's alcohol consumption, health advice received, and beliefs about the possible risks) and how much they moderate the relationship between previous consumption and alcohol intake during pregnancy.

## **2. Participants and Methods**

#### *2.1. Study Design*

A cross-sectional study was carried out, through interviews, on a representative sample of the pregnant women treated in a publicly managed university hospital in Seville (Spain). The sample was randomly selected from women who attended the morphology ultrasound clinic, located in the outpatient area of the hospital, in their 20th week of pregnancy, during a five-month period in 2016.

## *2.2. Data Collection and Participants*

The population of expectant mothers in the twentieth week of pregnancy in the health area of this university hospital during the period when the data were collected was 1664. The sample selection criteria were set as an interview with one out of every two pregnant women, to be chosen at random, i.e., 832 pregnant women. Of these, 426 agreed to be interviewed. The minimum desired sample size was 400 participants. All of them had the same gestational age, and in this regard, it was a homogeneous sample and one that was representative of the population of pregnant women treated by the aforementioned public hospital. For the collection of data, face-to-face interviews were conducted, carried out by health professionals who had previously been instructed on how to do so. The questionnaire was custom designed by the research group and was delivered under conditions that ensured the anonymity of those interviewed and the confidentiality of the information collected.

The eligibility criteria for inclusion in the study were: Pregnant women of 16 years of age or older, who speak and read Spanish fluently, who accepted and signed the informed consent for inclusion in the study. Further characteristics of the sample are described elsewhere [32].

## *2.3. Ethics*

Before the study was carried out, both its protocol and the questionnaire prepared by the research group were approved by the Clinical Research Ethics Committee of the University Hospital Virgen Macarena (Research code: ICG15/Internal code: 0254N-15).

As a prerequisite for conducting the interview, pregnant women were given oral and written information about the study and an informed consent form, which they had to complete and sign voluntarily if they wanted to be part of the study, delivering one copy to the research team and keeping the other. This documentation reflected all the information related to the objective of the study, as well as the guarantees of confidentiality, privacy, and preservation of anonymity in the responses. The participants gave their informed consent by signing and returning this form. The Helsinki declaration of 1975 and its subsequent amendments were respected.

#### *2.4. Questionnaire*

The instrument used for the collection of information was a questionnaire prepared and designed ad hoc for the study by the team of researchers who conceived and carried out this research project. Each of the completed questionnaires was given a code that preserved the anonymity of the users. The members of the research group included health professionals (a GP, two from the field of obstetrics, in particular—an obstetrician and a midwife—and a neonatologist), as well as professionals from the fields of psychology and sociology. The experience of all of them, as well as their knowledge in the field as a result of their research background, made it possible to prepare the customized questionnaire for the population to which it was addressed. Furthermore, a preliminary pilot was carried out in order to verify understanding of the questions, as well as the possibility of adding or removing categories in the answers to the multiple-choice questions. The questions regarding consumption patterns were taken from the Alcohol Use Disorders Identification Test (AUDIT) [33].

Most of the questions in the questionnaire provided the possibility of answering with several predetermined options; however, one of them was "other", which allowed the interviewer to note down all answers provided spontaneously by the pregnant women that were not in line with the categories established. Subsequently, the research group transcribed these answers to categorize them into the options that had been established beforehand, or put them into a new category, thus preventing information from being lost. In addition to these multiple-choice questions, there were also open-ended questions in the questionnaire that were recorded by taking notes. After the data were collected, categories were created for these answers based on a thematic analysis of them.

The questionnaire's content covers the following groups of variables:


perceived duration of damage resulting from alcohol consumption during pregnancy (categorized as seven possible answers: (1) During pregnancy; (2) during childhood; (3) first years; (4) many years; (5) lifelong; (6) she doesn't know; (7) other). The categorization of answers was performed following the piloting of the study, in which these same questions were asked, allowing those responding to provide open answers, which were subsequently categorized. Furthermore, the "other" answer option was provided, as described above.

