How many children live with persons with problematic drinking patterns across 19 European countries?

Abstract The harm from others’ drinking alcohol is well documented by researchers in different countries and with different cultural backgrounds across the world. Studies highlighted the use of alcohol in a problematic way as a crucial factor for a child’s normal physical, emotional, and social development. This study estimated the prevalence of children living with at least one person with a problematic drinking pattern (PPDP) in the household in 19 European countries. Furthermore, it examined age and gender differences in the prevalence of PPDP. Data from the Standardized European Alcohol Survey project (RARHA SEAS-1) and Eurostat was used to calculate the prevalence of children living with at least one PPDP. A four-item version of the Rapid Alcohol Problems Screen test was applied. Descriptive statistics and binary logistic regression analysis were conducted with SPSS. The prevalence of children living together with a PPDP varied from 6.2% in Italy to 35.2% in Lithuania. The total number of children living with PPDP was estimated to 9,271,150 in 19 European countries in 2015. The number of children living in an environment with PPDP is particularly high. Research in this area is important to bring this issue to policy agendas.


Introduction
The harm caused by the alcohol consumption of a household member is well documented by researchers in different countries and in different cultural backgrounds across the world Inoura et al., 2020;Jokinen et al., 2021). Children can experience different forms of harm from other people's drinking: harm to development (Andre et al., 2020;Fagerlund et al., 2011;Popova et al., 2017;Dejong et al., 2019;Tsang & Elliott, 2017); being left in unsupervised or unsafe situations, verbally or emotionally abused, physically hurt or neglected, exposed to domestic violence, witnessing verbal or physical conflict, drinking or other inappropriate behavior (Esser et al., 2016;Freisthler et al., 2020;Laslett et al., 2015). Children's exposure to problematic alcohol consumption contributes to their adverse childhood experiences (ACE) which can negatively affect their entire lives and can pass the ACE to future generations (Anda et al., 2002;Christoffersen & Soothill, 2003;Felitti et al., 1998;Flannigan et al., 2021;Galler & Rabinowitz, 2014;Haugland et al., 2021;Hughes et al., 2017;Kaplan et al., 2017;Szab o et al., 2021).
Harm to children caused by alcohol is often studied from a parental perspective (Velleman & Templeton, 2007). Maternal alcohol abuse can damage prenatal child development and cause fetal alcohol syndrome (FAS), which is a leading cause of intellectual disability, birth defects, and developmental disorders, yet is entirely preventable if mothers during the prenatal period do not use alcohol (Popova et al., 2017;Dejong et al., 2019;Tsang & Elliott, 2017). Children whose parents are diagnosed with alcohol use disorders have an increased risk of mental and behavioral disorders (Raitasalo et al., 2019), a risk of educational difficulties (Mangiavacchi & Piccoli, 2018), and criminality (O'Leary et al., 2020).
Problematic use of alcohol by caregivers is a significant factor in the inclusion of children in the protection system (Laslett, 2014;Tamutien _ e, 2018). In a study in the state of Victoria (Australia), Child Protection Services cases substantiated between 2001 and 2005 found likely alcohol abuse was identified in 33% of substantiations, 36% of protective interventions, and 42% of court orders. (Laslett et al., 2013). Tamutien _ e's (2018) study based on document content analysis of 203 case records in child protection in Kaunas (Lithuania) found that in 85.7% of the cases, the family had at least one problem drinker. Seay (2015) found the prevalence rates of parental substance use in child welfare across the eleven studies in the sample (US) ranged from 3.9% to 79% (M ¼ 35.04, SD ¼ 18.06). Cross-culturally in a diverse sample of nine countries (Australia, USA, Ireland, Chile, Thailand, Sri Lanka, Nigeria, Viet Nam, Lao PDR) about one in 25 children experienced severe alcohol-related harm by either being physically harmed or witnessing family violence in a 12-month period . However, parents, other caregivers, and any other family member with PPDP may have a negative impact on children (Berends et al., 2012;Dube et al., 2001;Orford et al., 2010). Therefore, the problem is very important and needs to be prevented. It is necessary to develop prevention and interventions to reduce alcohol-related harm to children. This requires knowing the scale of the problem.
Although Europe is the region with the highest alcohol consumption (World Health Organization, 2018), it is not known how many children live in households with PPDP. There were some studies on the prevalence of children living with PPDP in different countries. Some of them are focused on parental alcohol use disorders (Elgan & Leifman, 2013;Galligan & Comiskey, 2019;Velleman & Templeton, 2007) and are based on specific populations such as child protection and welfare clients (Tamutien _ e, 2018). There are no general numbers of the prevalence of how many children in Europe live within households with PPDP. Our study contributes to filling this gap.
The goal of this study is to estimate the prevalence of children living with at least one PPDP and to examine the age and gender differences in the prevalence of PPDP in 19 European countries (Austria, Bulgaria, Denmark, Estonia, Finland, France, Greece, Hungary, Iceland, Italy, Lithuania, Norway, Poland, Portugal, Romania, Spain, Sweden, UK) which participated in the Standardized European Alcohol Survey (Moskalewicz et al., 2016). Therefore, this is an important first step to evaluate how many children share a household with a PPDP. Although these numbers do not tell how many children suffer from alcohol-related harm, they are important and necessary to develop prevention and interventions to reduce alcohol-related harm to children and prevent alcohol-related adverse childhood experiences.

Data and sampling
Data from the Standardized European Alcohol Survey (SEAS-1) and Eurostat (2021) was used to calculate the prevalence of children living with at least one PPDP. SEAS-1 was implemented as part of the European 'Reducing Alcohol Related Harm' (RARHA) project in 2015/2016 (Moskalewicz et al., 2016). The survey was carried out in 19 EU/EEA countries representing different drinking cultures and survey traditions.
The definition of the target population was common for all surveys which sampled inhabitants in the country in the age range 18-64 even though some countries extended their sample to 65 years. Randomized sample selection was applied in all countries, but the sampling procedures were countryspecific. Most commonly, multistage stratified probability sampling was applied in 11 countries. In 7 countries the sample was drawn using a simple random design. One country used stratified random sampling and several used random sampling with a quota approach component. Half of the surveys applied computer-assisted telephone interviews (CATI), in seven surveys interviews were carried out face to face but computer-assisted (CAPI), and one survey applied to paper and pencil approach only. In a few countries, more than one model was applied. The total sample size was 32,576. In this study, we use a weighted sample size -32,266. Information in more detail about sample sizes in each country is provided in the SEAS Synthesis Report (Moskalewicz et al., 2016, p. 12-19).
All statistical analyses were conducted using probability weights in order to enhance the distribution fit between samples and population regarding gender and age group. By dividing the population proportion of each stratum by the empirical proportion of each stratum in the national datasets, a weighting variable was calculated. Information in more details about weighting is provided in the SEAS Synthesis Report (Moskalewicz et al., 2016, p. 18-19, p. 279-280).

Measures and methods
We have used Rapid Alcohol Problems Screen (RAPS) questions in our analysis. In the SEAS-1 survey, the RAPS questions as a brief instrument for screening purposes (Cherpitel, 2000) were included. The RAPS consists of four simple questions indicating alcohol mental disorders: Now I'd like to ask you some more questions about your drinking in the past 12 months: RAPS 1. Have you had a feeling of guilt or remorse after drinking? RAPS 2. Have you had a friend or family member tell you about things you said or did while you were drinking that you did not remember? RAPS 3. Have you failed to do what was normally expected from you because of drinking? RAPS 4. Do you sometimes take a drink in the morning when you first got up?
A total score ranges between 0 and 4. When used as a screening tool for alcohol dependence, at least one RAPS item must be answered positively. In this analysis, we used the term 'person with problematic drinking pattern' (PPDP) instead of 'alcohol dependent person' or 'person with alcohol mental disorder.' We argue that the answer 'Yes' to at least one RAPS question during the previous 12 months does not necessarily indicate an alcohol mental disorder. It is only a brief screening tool for dependence. To do a proper diagnostic conclusion about alcohol dependence, the Composite International Diagnostic Interview tool could be applied (CIDI) (Robins et al., 1988). This survey also includes 16 CIDI questions. However, CIDI questions were optional for only 6 countries in the survey. Thus, we decided to use RAPS questions, but to use a more discreet term 'person with problematic drinking pattern.' Furthermore, to identify people with more problematic drinking patterns we added another screening variable with at least two RAPS positive answers.
In the SEAS-1 survey, each respondent was asked the question 'SD_7. How many persons living in your household are younger than 18 years old?' In our study, a child is a person under 18 years old. We used this question to identify the number of children who live with a PPDP. Furthermore, we calculated the prevalence of children living with PPDP and the total number of children living with a PPDP in 19 European countries. The total number of children in 19 European countries on 1 January 2015 was downloaded from Eurostat Database, online data code: DEMO_PJAN (Eurostat, 2021).
The equation of the total number of children living with PPDP in each country could be expressed as follows: NCPtotal number of children living with PPDP. PCPprevalence of children living with PPDP (data source: SEAS-1). NCtotal number of children living in each country (data source: Eurostat).
The total number of children living with PPDP in all 19 European countries was estimated by summing up the number of children living with PPDP in each country.

Country grouping
In the analysis, we have divided countries regarding their geographical position and drinking culture into three major groups: Northern Europe (Lithuania, Iceland, Sweden, UK, Estonia, Finland, Norway, and Denmark), Central Europe (Bulgaria, Poland, Austria, Hungary, Romania) and Mediterranean Europe (Croatia, Spain, Greece, France, Portugal, and Italy). The division was based on the results of SEAS-1, published by Moskalewicz et al. (2016). We choose to mean alcohol consumption per drinking day as an indicator of health-damaging drinking culture. This indicator shows a clear north-south gradient. Northern European countries showed the highest volume of alcohol consumption while Mediterranean countries showed the lowest alcohol consumption volume per drinking day (Moskalewicz et al., 2016, p. 106-107).
We applied descriptive statistics analysis and binary logistic regression analysis in our study. The aim of the binary logistic analysis was to assess associations between the prevalence of PPDP and the main demographic characteristics of respondentsage and gender. The significance level was set at p < 0.05. All analysis was conducted with SPSS 26.0. Cases with missing values were not included in the analysis.
The study follows the principles of the Declaration of Helsinki.

Results
Results of our study show that 19.2% of all respondents answered 'Yes' to at least one RAPS question and 8.3% answered 'Yes' to at least two RAPS questions in 19 European countries (see Table 1). The proportion of positive RAPS answers was lower among those living in households with children when compared with the whole sample. We found that 16.8% of those living with children answered 'Yes' to at least one RAPS question and 6.9% answered 'Yes' to at least two RAPS questions.
In households with children, the proportion of at least one positive RAPS answer varied from 5.5% in Italy to 35.2% in Lithuania (see Figure 1). The highest proportion of at least one positive answer to RAPS questions was among the Northern European countries, while the lowest frequency was recorded in countries of Mediterranean Europe. Central European countries Bulgaria (19.1%), and Poland (16.5%), reported results closer to Northern countries like Denmark (17.5%) and Norway (19.5%), then to other countries in the Central European group. The countries with the lowest proportion of positive answers to at least RAPS questions were Italy (5.5%), Portugal (6.6%), Romania (7.8%), and Hungary (8.6%). The pattern of the proportion of at least two positive answers to RAPS questions was very similar to the pattern of the proportion of at least one positive answer to RAPS question among 19 European countries. For more detailed data, see Supplementary Appendix 1.
The average number of children living in households with PPDP who answered 'Yes' to at least one RAPS question varies a lot across 19 European countries (see Table 2). The lowest prevalence of children living in a household with the PPDP is in Italy (6.2%), and the highest is in Lithuania (35.2%). Less than 10% of children live with PPDP in France, Greece, Hungary, Italy, Portugal, and Romania. More than 20% of children living with PPDP are in North European countries: Estonia, Iceland, Lithuania, Sweden, and the UK. The total number of children sharing the household with PPDP who answered 'Yes' to at least one RAPS question was estimated to 9,271,150 in 19 European countries. The number of children living with PDPP who answered 'Yes' to at least two RAPS questions was lower (3,659,691), but the pattern of prevalence remained similar across different countries. Children living in such households were even more likely to face harm from others drinking.
The prevalence of respondents living with children and reporting problematic drinking patterns (at least one positive answer to RAPS questions) was analyzed by gender and age group. Table 3 shows the results of binary logistic regression analysis. We found that the prevalence of respondents living with children and reporting problematic drinking patterns OR differences were statistically significant (p < 0.05) between males and females in 17 out of 19 European countries. The results show that men living with children had a higher problematic drinking risk when compared to women living with children. The only exceptions were Estonia and France which did not show a significant difference. The results of regression analysis indicated that the maximal relative gap between men and women varied from 1.60 in the UK, to 24.72 in Romania. The odds ratio gender gap was the largest among countries with relatively low rates of the prevalence of women living with children and reporting problematic drinking patterns. For detailed results see Supplementary Appendix 2.
Differences between the prevalence of respondents living with children and reporting problematic drinking patterns OR in different age groups were statistically insignificant (p > 0.05) in 14 out of 19 countries. Only five countries from Northern Europe (Estonia, Denmark, Iceland, Norway, and the UK) showed statistically significant OR differences between age groups. The prevalence of respondents living with children and reporting problematic drinking patterns OR were higher among the age group 18-34 when compared to the age group 50-65 in all five countries. For more detailed data, see Supplementary Appendix 3.

Discussion and conclusions
The goal of this study was to examine the prevalence of children living with at least one PPDP in the household in 19 European countries. We found that the prevalence of children living together with a PPDP varied from 6.2% in Italy to 35.2% in Lithuania. The total number of children living with PPDP was estimated to 9,271,150 in 19 European countries in 2015. Differences in the prevalence of children living with PPDP were identified between different European countries. This could be explained by the drinking practices identified by Kilian et al. (2021). Comparing country-specific drinking practices classes (Kilian et al., 2021(Kilian et al., , p. 2019(Kilian et al., -2021 with our estimates of the prevalence of children living in a household with PPDP, one can see similar trends. Countries, where we have identified a lower percentage of children living with PPDP, have a higher prevalence of light to moderate drinking patterns. Conversely, countries with a higher percentage of children living with PPDP have a higher prevalence of drinking patterns characterized as infrequent heavy drinking and regular drinking. Our study contributes to research revealing that using alcohol in a problematic way is less prevalent in households with children than those without (Kavanagh et al., 2011;McKetta & Keyes, 2019;Paradis, 2011). However, the study also reveals that almost 9.3 million children have lived in households where at least one adult may have problematic drinking patterns in the past year in 19 European countries. It should be noted that the RAPS method we applied is used to screen alcohol use disorders and is a sensitive measure of alcohol dependence (Bloomfield et al., 2013;Cherpitel, 2000Cherpitel, , 2002Cherpitel et al., 2005). We do not claim that a positive RAPS is a diagnosis of alcohol dependence but that it reveals an indication of having alcohol problems, and more detailed testing would be needed to confirm the diagnosis.
Men exceeded women in PPDP in households with children, and this was statistically significant (p < 0.05). Our findings confirm the research trend that gender differences in alcohol consumption remain universal, although the sizes of gender differences vary Knibbe & Bloomfield, 2001;Wilsnack et al., 2009). Our results are in line with M€ akel€ a et al. (2006), where the men and women drinking patterns in Nordic countries are closer to each other than elsewhere. In addition, we found that the drinking pattern in Bulgaria is close to the Nordic countries.   Note. Data sources: RARHA SAS-1 and Eurostat. All data came from year 2015. Wilsnack et al. (2009) found that the association of youth with heavy episodic drinking may be primarily an Anglo-European pattern. When we look at all of 19 European countries, our odds ratios for the prevalence of respondents living with children and reporting problematic drinking pattern by age group differs across countries. In 17 of the 19 countries, there is a trend for people aged 18-34 in a household with children to have a PPDP. However, these differences are statistically significant only in 5 Northern Europe countries: Estonia, Denmark, Iceland, Norway, and the UK. In Lithuania, which has the highest number of children living in households with PPDP, the 50-64 years' age group tends to prevail over the younger, but this result is not statistically significant. Comparing the results of our study in 5 Nordic countries with the study of Kilian et al. (2021), we see that statistically significantly more young people with a problematic drinking pattern have the lowest prevalence of drinking practices in the category 'Light to moderate drinking without risky singleoccasion drinking.' This situation raises the question of alcohol-related harm to children and its prevention.
In countries where similar studies were done, a large proportion of children are exposed to problematic alcohol use or alcohol-related harm. Due to different measurements, it is still difficult to compare the overall prevalence in different countries, but we could capture the tendencies. In Ireland, an estimated 14-37% of children are possibly impacted by parental alcohol dependency (Galligan & Comiskey, 2019). In Sweden, the proportion of adolescents classified as having parents with alcohol problems was 20.1%, and 44.0% reported that they think someone close to them drinks too much alcohol and 9.6% that this has hurt them or caused them problems (Elgan & Leifman, 2013). The Health Survey for England and General Household Survey generated consistent estimatesaround 30% of children under 16 years (3.3-3.5 million) in the UK lived with at least one binge-drinking parent (Manning et al., 2009).
Studies show that children living in households with PPDP experience alcohol-related harm all over the world (Caetano et al., 2019;Florenzano et al., 2015;Kaplan et al., 2017;Laslett et al., 2015). Although our study does not show how many children suffer from alcohol harm, the results suggest that a large proportion of children in Europe could be at risk of experiencing an ACE due to PPDP within the household. That could be detrimental to their present lives and will have negative consequences for their lives in adulthood (Anda et al., 2002;Koponen et al., 2020;Olsson et al., 2019;Raitasalo et al., 2019;Raitasalo & Holmila, 2017).
Early and ongoing screening for alcohol use by mothers and fathers could help identify individuals at risk for problem drinking and alcohol-related problems during adolescence and young adulthood (Parra et al., 2020). However, according to the ACE study (Felitti et al., 1998), problematic alcohol use by a household member (not necessarily the mother or father) can contribute to a harmful childhood. From our study, we cannot draw any conclusions about the causality of ACE and children living in the household with PPDP, but there could be such a risk, and prevention and intervention should be planned and implemented.
Our study has few limitations which must be mentioned. First of all, there are several methodological limitations. It should be noted that in different countries the ways that data was collected were different. Response rates and sampling procedures varied in different countries a lot. This problem may affect the findings of this study. However, all surveys were representative, and data were weighted to avoid bias. Data sampling and response rates are briefly described in SEAS Synthesis Report (Moskalewicz et al., 2016, p. 12-19). Furthermore, some social groups, such as homeless Table 3. Odds ratios for the prevalence of respondents living with children and reporting problematic drinking pattern (at least one positive answer to RAPS questions) and their 95% confidence intervals (CI) by gender and age group in 19 European countries. people and those living in institutions, could have been not represented in this survey. Another limitation is that the findings of our study do not allow us to draw conclusions about the children's situation during the whole childhood. The RAPS items in RARHA SEAS-1 survey only cover the period of the last 12 months. Furthermore, we have not examined the frequency of rapid alcohol problems' episodes. It is likely that the actual number of children living with PPDP may be higher due to the under-reporting of alcohol use. Heavy drinking and non-routine drinking patterns may be associated with greater under-reporting of alcohol consumption. Self-reported alcohol consumption typically amounts to 40-60% of total alcohol sales (Boniface et al., 2014, p. 2).
Another limitation is that we do not have any information regarding children's sociodemographic characteristics such as gender, age, school attendance, etc. Moreover, we do not know if children face any harm related to others' drinking. More detailed research is needed to answer the question of how many children, at what age, what harm they experience when living with PPDP, who is the person with a PDP (mother, father, mother's/father's partner, grandparent, or other siblings), and how many household members have a PDP. This information could be crucial for planning prevention and interventions.
Our study contributes to the research field of studies on the prevalence of children who live in households with PPDP, whether they are parents or other adults. To our knowledge, this is the first such kind of study at the European level. Knowing the prevalence of children living within households with PPDP is very important in assessing the scale of the problem and in planning prevention measures beyond the national level alone.
The fact that around 9.3 million European children under the age of 18 have lived in a household in the last year with a PPDP or even potentially alcohol-dependent adult is a serious public health and social problem that needs to be addressed. This calls for further strategies to prevent child exposure to problem drinking environments. A child-friendly, nurturing environment is incompatible with problematic alcohol consumption by adults. In European countries, the number of children living in a PPDP environment is particularly high. It cannot be ignored if we want to ensure the wellbeing of the current generation of children and future generations at the national and regional levels. Research in this area is important to bring this issue to policy agendas.