Family Structure through the Adolescent Eyes: A Comparative Study of Current Status and Time Trends over Three Decades of HBSC Study

Abstract

This study aimed to investigate how family structure varies and identify its time trends in European and North American countries using data from seven surveys conducted between 1994 and 2018 according to the Health Behaviour in School-aged Children (HBSC) study. The current family structure in 44 countries was described and time trend analysis of 28 countries was performed. Adolescents were asked whom they lived with in their home to describe family structures. Family structures showed distinct patterns and dynamics between countries. In 2018, in all countries, 73% of adolescents lived with both their mother and father; 14% and 5% of adolescents lived in a single-parent family and stepfamily, respectively; and around 9% of adolescents lived in another family type. In the period 1994–2018, the proportion of young people living in intact families decreased from 79.6% to 70.0%, on average about 10 percentage points. There were no significant changes in the prevalence of single-parent families and stepfamilies, but a significant increase in the number of adolescents living without either parent was revealed. The findings have implications for cross-national adjustment of adolescent health, well-being, and behaviours, and for critical analysis of socioeconomic family resources.

1. Introduction

The family is the primary place for the development and socialization of adolescents. In the family, they learn the norms of behaviour and the values of life. Many factors determine the functioning of the family and its significance for young people’s development, but the structure of the family is undoubtedly of great importance [1].

Different family structures include nuclear families, single-parent families, stepfamilies, families headed by two unmarried partners (either of the opposite sex or the same sex), adoptive families, extended families, and grandparent families [2]. Nuclear families are usually married couples who have any number of children. Although there are differences in the definitions among observers, in this study, we use the intact family structure equivalent to the nuclear family structure as a concept to refer to those children who are living in families in which both biological parents are present in the household [3]. This type of family structure contrasts with the non-intact family. This term is also known as the reconstructed, reconstituted, or blended family in which one of the partners is not a biological parent [2]. From the child’s perspective, the non-intact family type includes children living in a single-parent family, stepfamily, or living with someone or somewhere else. Parental separation or divorce is the most common cause of family composition disintegration [1,4,5,6].

The literature on the impact of family structure on child development and well-being is huge. It covers several disciplines, including economics, demography, sociology, and psychology. A recent scoping review evaluated 283 studies focused on the associations between family structure and well-being [7]. Nearly three decades of research on the impact of changing family structures shows that the highest levels of physical, emotional, and academic well-being are achieved among children living in nuclear or intact families, controlling for other family circumstances [4,8,9,10,11,12,13].

Research shows that living with both parents constitutes a protective factor for normative development while living with a single parent or in stepfamilies has previously been associated with a higher risk of poor psychological adjustment (depression, anxiety, poor perception of well-being, and social isolation) [14,15,16,17]. Young people who are living in non-intact families are also more likely to engage in risky behaviour, such as sexual initiation [18]. The most disturbing finding that researchers have found is the relationship between impaired family structure and suicide [19,20]. Changes in family structure can also predict poorer behavioural outcomes such as cigarette smoking and/or use of alcohol in adolescence [21,22,23]. Living in a single-parent family or in a reconstituted family was unfavourably associated with physical activity, sport participation, and screen-based behaviours among youth [24]. Adolescents from non-intact families had a less healthy diet [25,26] and higher risk of being overweight [27,28].

Research on the impact of changing family structures on general life satisfaction found that children living with both biological parents reported higher life satisfaction than children living with a single parent or in stepfamilies [29,30]. However, the effect size of such a parenthood/family structure on child well-being varied between countries, with no systematic link to differences in policies [10,31,32]. Evidence has shown that in adolescence, the quality of family relationships might be more important for health outcomes than family structure per se [33,34,35]. Shared physical custody after divorce is an increasing pattern in some countries such as Sweden [36]. In a study of 37 countries [37], it was found that adolescents’ life satisfaction in such an arrangement is higher than in asymmetric arrangements, although this is related to the children and family characteristics. Moreover, a large body of research has shown that families can function well and children can thrive in a variety of family structures if families are stable, have good parent–child relationships, and are financially secure [38,39].

Despite the examples of the strong implications of family structure on child development, data on the prevalence of different types of families can only be found mostly in official statistics. For many countries, official statistics contain only data on the frequency of parental divorces, families by number of children, births to unmarried women, and other data that are successfully recorded during the census [40,41,42,43].

It should be noted that family structure can be measured in two different ways: from a family-centred perspective and from a child-centred perspective [32]. From the first perspective, generally for national databases, information is gathered in population surveys assessing the proportion of some types of family structure in all families with children. The sampling frame for the surveys is most typically a household address-based one. The person responding to questions regarding the family situation is an adult, rather than a dependent child. From the second child-centred perspective, the proportion of all children who are living in some type of family structure is estimated. Researchers record the child’s response and use the school rather the household address as a sampling frame. In this way, the questions on family structure are measured by the responses of the children, not the adults. Adult and child perceptions of family structure, of course, may differ. The perceptions of the children are certainly more important when examining the effects of family structure on children’s health. However, very few cross-national studies collect information on family structure from the child’s perspective [44].

The Health Behaviour in School-aged Children (HBSC) study is a World Health Organization (EURO) collaborative cross-national project that aims to investigate the health and well-being of adolescents in Europe and North America (see: www.hbsc.org (accessed on 22 April 2022)). Cross-sectional surveys with school-aged children of 11, 13, and 15 years of age have been conducted regularly since 1983/84 across an increasing number of countries [45,46]. As HBSC focuses on the social context of young people’s health, health behaviour, well-being, and education, family measures are undoubtedly of great importance. The determination of family structure is one of the key measurements used since the first HBSC survey. The family structure is assessed based on the child’s own responses, and the methodology for collecting this information has not changed since 1994. The impact of family structure on adolescent health, well-being, and health behaviour has already been highlighted in a number of papers from the HBSC survey data. The concept of family has been used in more than 100 HBSC scientific papers (see: www.hbsc.org/publications/ (accessed on 1 April 2022)). Comparative studies of family structure have been conducted by Chapple in 2009 [32] and Låftman in 2010 [10], but these studies, despite their age, were limited to single-mother families. Tables and/or graphs showing family structure by HBSC countries have been presented in the international reports for each HBSC survey [47,48,49,50,51,52,53]; however, no detailed analysis of family structure time trends using data of the HBSC study has been performed to date.

Consequently, using the large datasets from HBSC surveys, the present study aimed to answer the following two questions: (1) how does family structure vary between countries participating in the HBSC, and (2) what time trends in family structure are shown over the three decades of HBSC study?

2. Materials and Methods

2.1. Study Population

In this study, we used data from the HBSC surveys that were conducted in the period of 1993/1994 (coded as 1994) to 2017/2018 (coded as 2018), which spanned almost three decades. During this period, seven surveys were conducted every four years. The number of countries or regions participating in the study increased from 24 (in 1994 [47]) to 45 (in 2018 [53]) (hereinafter, the Flemish- and French-speaking regions of Belgium, and the England, Scotland, and Wales regions of the United Kingdom were considered individually as countries). In all surveys, data collection followed the standard methodology outlined in the HBSC protocols and included country-representative samples of 11-, 13-, and 15-year-old students. The data were gathered through a self-filled questionnaire administered in schools. A detailed description of the aims, theoretical framework, and survey methodology of the HBSC study can be found in the study website (www.hbsc.org) (accessed on 22 April 2022), international protocols [54], and international reports, which summarized the results of each of the seven surveys [47,48,49,50,51,52,53]. The data were obtained from the International HBSC Data Bank (Bergen University, Bergen, Norway).

The status of the family structure was described, using data from the 45 countries that participated in the HBSC survey in 2018. Since no data on family structure was received from Azerbaijan, the final list of countries in the present analysis included 44 countries, with a total of 215,875 surveyed students with non-missing responses on family structure.

Time trend analysis was performed using the data of 28 countries in which the survey was conducted in 1994 (or at least in 1998) and in 2018 (or at least in 2014). Overall, 906,235 students of both sexes aged 11, 13, and 15 years were included in this analysis: 94,618 students in 1994; 116,173 students in 1998; 128,756 students in 2002; 141,630 students in 2006; 136,573 students in 2010; 148,466 students in 2014; and 140,019 students in the 2018 study waves.

2.2. Ethics

This study conformed with the principles outlined in the World Medical Association Declaration of Helsinki. National teams obtained ethical consent from the institutional ethics committee(s), when required. Parental consent was passive in most countries. Students were informed orally and in writing that participation in HBSC was voluntary. Students did not provide any personal information (such as name, classroom, teacher) to guarantee the students’ anonymity and to ensure the confidentiality of the personal responses.

2.3. Measurements

Family structure. The measure of family structure aimed to gain a full picture of the (sometimes complex) type of families in which children lived [54]. It remained unchanged and mandatory over the study period (1994 to 2018), except for editing the preamble of the question. The measure examined the family composition in the main home where the child lived all the time or most of the time. Following the suggestion of some countries, in the 2002, 2006, and 2010 surveys, the measurement of family structure was supplemented by an analogous examination of the family composition in the second home, if any (for instance, among children whose parents were divorced), but such data were not included in the analysis of the present study. The validity of the question was confirmed in all HBSC surveys, with a strong influence of family structure on adolescent lifestyle and health [47,48,49,50,51,52,53].

To define the type of family structure, the students were asked to answer a question of whom they lived with. When choosing the answer “yes” or “no”, they had to indicate whether they lived with their “mother”, “father”, “stepmother (or father’s girlfriend)”, “stepfather (or mother’s boyfriend)”, “grandmother”, “grandfather”, “I live in a foster home or children’s home”, or “I live with someone or somewhere else: please write it down”. The students who responded that they lived with their “mother” and “father” were defined as living with two biological parents or living in an intact family. Those who responded that they lived with their “mother” but not their “father” or “stepfather” were defined as living with their mother only (single-mother family). In addition to these types of family structure, four more types were formed, as shown in Figure 1, with a total of six types. The group of adolescents living in an “other type” of family structure included those who reported living in a foster home or children’s home, living with someone or somewhere else, living with grandparents, or living in a homosexual family.

Reasons for not living with both parents. This measure was developed with the aim of identifying why the teen did not live with both parents. This optional measure has been tested by several countries participating in HBSC. At first, students were asked if they lived with both parents. If the answer was no, then they were asked why. The answer options were: 1—parents are divorced; 2—mother/father is dead; 3—mother/father is living/working far away; 4—never seen father/mother/parents; 5—other reasons.

2.4. Statistical Analysis

The proportions of family structure types were estimated from the HBSC data of each country by the wave of the survey. The aggregated data of all countries was estimated by weighting data by the country’s sample size to ensure that the sample was representative of the general population. In the time trend analysis, a z-test with Bonferroni correction was conducted to estimate the significance of the proportion changes. Associations between variables were assessed using the non-parametric Spearman correlation. A significance level was set at p < 0.05. Descriptive analyses were performed with SPSS (version 21.0; SPSS Inc., Chicago, IL, USA, 2012).

3. Results

3.1. Family Structure in HBSC Countries in 2018

Based on data that was aggregated and weighted by country sample, the percentages of adolescents who reported living in a certain type of family were estimated (Figure 2). In contrast to crude data (69.7%, see Figure 1), the weighted data showed that 72.7% of young people reported living in an intact family, i.e., in a family with both parents. Fourteen percent of respondents reported that they lived with only one parent, but most of them lived with their mother. About five percent of the respondents indicated living in stepfamilies, mostly in families with the biological mother and stepfather. Eight percent of adolescents lived in the other type of household, such as being cared for by grandparents or the state (in foster homes or children’s homes, for instance).

There was much variability between countries in the proportion of adolescents who reported living with both parents (Figure 3). In 8 of 44 countries, over 80% of the adolescents lived with both parents (Switzerland (80.1%), Greece (81.4%), Slovenia (81.7%), Croatia (84.2%), Georgia (85.1%), North Macedonia (89.1%), Armenia (89.8%), Albania (90.5%)) compared to adolescents from Greenland, of whom less than half (48.4%) reported living in an intact family.

There was also high variability between countries regarding other family structure types. For instance, Romania, Kazakhstan, and Scotland excelled as countries where more than 20% of adolescents indicated that they lived in a single-mother family. In contrast, the lowest proportions of children living in single-mother families were registered in the Netherlands (4.7%) and Albania (5.1%) where the corresponding figures were about five times lower than in Romania (24.9%) (Figure 4). The percentage of adolescents living in families with their mother and stepfather ranged from under 0.2% in Albania, Armenia, and Georgia to over 8% in Wales and Hungary. In North Macedonia, Armenia, and Georgia, very few (less than 2%) of the young people responded that they lived with someone other than at least one parent. However, in Greenland, England, Norway, and Estonia, this proportion reached 20% (data for the last two indices are not presented).

3.2. Reasons for a Non-Intact Family Structure

Information on the reasons for the change in the family composition was collected through an optional question in the HBSC survey questionnaire. In 2014, the question was tested in Lithuania, and in 2018, it was adopted in Armenia, Moldova, and Ukraine.

Table 1. Reasons given for not living with both parents in several countries.

Country	Year	N	Parents Are Divorced	Mother/Father Is Dead	Mother/Father Is Living/Working Far Away	Never Seen Father/Mother	Other Reasons
Lithuania	2014	1875	58.6	14.9	13.4	3.9	9.2
Armenia	2018	647	27.4	16.8	22.9	8.0	24.9
Moldova	2018	1078	50.0	12.1	18.6	3.6	15.7
Ukraine	2018	1762	63.4	16.6	3.1	4.7	12.2

shows the distribution of the reasons mentioned by the children for not living with both parents. The most common cause was parental divorce, the frequency of which ranged from 27.4% in Armenia to 63.4% in Ukraine. The proportions of families with a deceased parent in the selected countries were almost equal (about 15%) while in Lithuania, Armenia, and Moldova, the proportion of families where the mother/father was living or working far away was noticeable (13.4% to 22.9%).

3.3. Time Trend of Family Structures over 1994–2018

Time trend analysis of family structures was performed using data from 28 countries in which the HBSC surveys were conducted in 1994 (or 1998) and 2018 (or 2014). The data presented in

Table 2. Comparison of the family structure between surveys in 1994 (or 1998) and 2018 (or 2014) across 28 HBSC countries.

Country or Region	Survey in 1994 (or 1998)						Survey in 2018 (or 2014)
	Year	N	Both Parents	Single Mother	Mother and Stepfather	Other	Year	N	Both Parents	Single Mother	Mother and Stepfather	Other
Austria	1994	5349	85.2	8.0	3.0	1.8	2018	4032	71.7 *	13.5 *	2.7	10.2 *
Belgium (Flemish)	1994	4476	86.0	6.8	4.0	0.6	2018	4034	72.3 *	15.1 *	4.0	4.8 *
Belgium (French)	1994	5196	75.5	10.3	6.3	4.8	2018	4020	65.3 *	14.0 *	5.3	11.9 *
Canada	1994	6699	77.1	11.8	6.3	0.8	2018	12707	71.3 *	14.6 *	5.0 *	5.2 *
Czech Republic	1994	3584	83.8	8.1	5.7	0.7	2018	11070	71.3 *	9.6 *	4.7 *	12.7 *
Denmark	1994	3867	74.8	12.6	8.2	1.1	2018	3073	76.7	6.3 *	1.2 *	14.0 *
England	1998	6373	68.2	14.4	11.0	2.9	2018	3397	58.8 *	14.1	3.8 *	21.2 *
Estonia	1994	3516	76.2	15.1	6.1	1.1	2018	4725	67.0 *	8.1 *	3.9 *	19.4 *
Finland	1994	4181	76.6	12.4	5.8	1.9	2018	3100	78.1	8.8 *	5.8	4.5 *
France	1994	3971	81.6	10.0	4.9	1.0	2018	9170	66.0 *	13.5 *	5.8 *	11.4 *
Germany	1998	4770	81.1	9.7	5.3	1.4	2018	4061	73.9 *	13.1 *	5.1	5.8 *
Greece	1998	4285	91.1	5.6	1.0	1.4	2018	3863	81.4 *	10.3 *	1.4	4.9 *
Greenland	1994	1238	57.7	17.9	7.3	11.6	2018	1168	48.4 *	17.6	6.2	21.6 *
Hungary	1994	5775	78.5	10.8	5.7	2.9	2018	3748	70.9 *	12.5 *	8.2 *	5.5 *
Ireland	1998	4352	86.1	8.1	0.9	3.2	2018	3833	72.3 *	14.0 *	4.1 *	7.9 *
Israel	1994	4103	91.5	5.5	1.4	1.0	2014	6104	86.1 *	9.8 *	1.8	0.9
Latvia	1994	3774	74.1	15.6	6.7	2.1	2018	4412	62.3 *	14.8	4.3 *	15.7 *
Lithuania	1994	5425	82.7	10.9	4.4	1.0	2018	3797	68.2 *	11.1	5.6 *	13.6 *
Norway	1994	4920	78.3	10.8	6.6	0.9	2018	3127	61.2 *	11.4	4.5 *	19.3 *
Poland	1994	4519	88.4	7.1	2.2	1.1	2018	5164	78.0 *	8.8 *	1.7	10.2 *
Portugal	1998	3720	85.4	7.9	2.6	2.5	2018	5655	70.6 *	10.4 *	3.3	13.8 *
Russian Federation	1994	4001	75.7	16.6	5.2	1.6	2018	4260	69.9 *	8.3 *	5.0	15.9 *
Scotland	1994	4770	75.8	14.6	4.9	1.7	2018	4773	65.0 *	20.2 *	5.2	7.3 *
Slovakia	1994	3402	88.5	7.6	1.9	0.6	2018	3911	71.4 *	12.2 *	4.3 *	9.9 *
Spain	1994	4431	90.8	6.2	1.3	0.7	2018	4320	78.8 *	11.8 *	2.2 *	5.6 *
Sweden	1994	3570	76.0	11.3	6.5	2.4	2018	4185	71.6 *	12.5	5.1 *	6.5 *
Switzerland	1998	3997	83.3	10.4	3.6	1.1	2018	7375	80.1 *	7.1 *	3.6	7.9 *
Wales	1994	3851	77.2	12.3	6.3	1.3	2018	13039	66.7 *	18.7 *	8.2 *	3.3 *

Notes: Proportions of adolescents who were living with a single father and living with a father and stepmother are not presented because they have only a small proportion in the family structure; * p < 0.05 compared with the data of the survey conducted in 1994 (or 1998).

allow a comparison of the proportions of all types of family structures between the selected survey waves by countries. In all countries, except Denmark and Finland, the proportion of intact families significantly decreased. There was a small change in the rank of countries that were sorted by the proportion of intact families, as the proportions of intact families that were estimated in 1994 (or 1998) and 2018 (or 2014) within selected countries were highly correlated (r = 0.79). In 17 of 28 countries, the proportion of adolescents living in a single-mother significantly increased (the highest increase was in Belgium (Flemish) and more than doubled to 15.1%) while in the rest of the countries, the percentage of children living in single-mother families decreased or did not change significantly.

In all countries, the proportion of the other type of family structure (adolescents living with someone other than at least one parent) significantly increased. On average, the share of such families reached 9.3% while in several countries, such as Greenland and England, that figure was over 20%. Compared with the data of the 1994 HBSC survey, the highest increases were observed in Norway (18.4%), Estonia (18.3%), Russia (14.3%), Latvia (13.6%), and Lithuania (12.6%).

Figure 5A shows the time trend of the proportion of intact families over 1994–2018 in the 28 countries. In this period, the proportion of families where the adolescent lived with both parents decreased from 79.6% to 70.0%; thus, the overall decrease was almost 10 percentage points. Similarly, Figure 5B shows the time trend of the proportion of families where the adolescent lived with a single mother. Here, it is difficult to notice any regular change trend. The mean percentage of this type of family structure during the observation period was about 12%.

4. Discussion

The present study aimed to answer to the following two questions: (1) how do family structures vary between countries, and (2) what are the time trends shown in family structures over the three decades of the HBSC study? Extensive data from the HBSC study conducted between 1994 and 2018 was used to answer these questions. Over this period, in seven waves of the HBSC survey, adolescents from many countries reported on the composition of their family by responding to a question regarding whom they lived with. In contrast with the data from the official statistics, the results of this study are visible “through the eyes of adolescents”, so they reflect the child’s perspective in regard to family structure. Thus, the present study was conducted in accordance with the UN Convention on the Rights of the Child, which states that “Children have the right to give their opinions freely on issues that affect them. Adults should listen and take children seriously” [55].

Overall, the results of the present study revealed a wide variety of family structures in the European region and North American countries. The results from the analysis outlined significant shifts in the composition of families over the last three decades. These results confirmed the basic idea that the family is not a closed and static unit but a complex and dynamic system, both affecting and being affected by social, cultural, and historical development [1,5,56] and by individual cycles and life course transitions [57,58]. Adolescence is one of the life course periods that is most sensitive to family structure change.

Using aggregated data from all countries, we found that in 2018, around 73% of adolescents lived with both their mother and father, i.e., in an intact family; 14% and 5% of young people lived in a single-parent family or a stepfamily, respectively; and around 9% lived in the other types of family. According to the data from the countries that completed the HBSC surveys, in almost all waves, the proportion of adolescents who reported living in intact families decreased by about 10 percentage points compared to the initial wave of the survey conducted in 1994. Although there were no significant changes in the prevalence of single-parent families and stepfamilies during the follow-up period, a significant increase was revealed in the number of adolescents living elsewhere or with someone other than their biological parents. However, variation between countries was markedly contrastive.

Two-parent families constitute the majority of families with children around the globe, but every region of the world has distinct patterns regarding family structures [41]. Our findings are in line with the literature [41] showing that intact families are more common in Eastern European countries while Western countries have higher rates of adolescents living in non-intact families (and subsequently lower rates of adolescents living with two parents), though there is strong variation among countries. Different cultural and societal norms and political and economic factors can account for many of these differences [6,41,59]. In particular, the countries with the highest rate of intact families or the fewest rates of single-parent families or stepfamilies often have strong religious traditions and strong familistic cultures [42].

In the present study, the time trends of the family structure types were also studied from the child’s perspective. The HBSC data from 28 countries showed that during the period 1994–2018, the proportion of adolescents who reported living in intact families decreased from 79.6% to 70.0% on average, although there were no significant changes in the prevalence of single-parent families and stepfamilies during the follow-up period. Although there may be differences in the definitions of the types of family structures, it can be observed that this finding correlates with data reported by the OECD Family Database, which states that on average across the OECD countries, the proportion of children living with two married parents decreased between 2005 and 2018—from 73% to 66%—while the share of children living in households with a single-parent remained stable [43]. This phenomenon is related to increasing trends of divorce or separation [1]. The optional question regarding the reasons adolescents did not live with both parents, which was utilized in the surveys of several HBSC countries, confirmed that parental divorce was the main reason for over half of the cases of children living in a non-intact family in Lithuania, Moldova, and Ukraine. Demographers are increasingly interested in another family phenomenon—the growing prevalence of cohabitation—which can significantly reduce the percentage of intact families. According to the OECD Family Database [43], living with two cohabiting parents is becoming increasingly common in Belgium, the Czech Republic, and Poland, where in each case, the proportion of children living with two cohabiting parents increased by over 10 percentage points between 2007 and 2018. The HBSC study did not investigate this phenomenon, but it may partly explain the low proportion of intact families in several countries, which is a result of the diffusion of cohabitations.

The results from the present study confirmed the painful social fact that a large proportion of adolescents live in single-parent families, most of which are headed by single mothers. In 2018, it was recorded that 12% of children only lived with their mother and 2% only lived with their father. This figure is in line with official statistics; according to the Eurostat data, in 2020, approximately 14% were single-parent households with dependent children [60]. According to the 2020 U.S. Census Bureau [61], out of about 11 million single-parent families with children under the age of 18, 80% were headed by single mothers; nearly one-third of them live in poverty. The percentage of children living in single-mother families is considered an indicator of child poverty [40]. There are several reasons for this. Recent demographic data show that, for example, in the USA, around half (49.5%) of single mothers have never married, almost one-third (29.9%) are divorced, and 20.5% are either separated or widowed [61]. The share of single-mother families varied considerably from one country to another. To compare the observed variation with data from other sources, we sorted the countries by the values of the proportion of single-mother families and compared their rank with analogous ranks found in official statistics. There was a significant correlation (ρ = 0.449, p = 0.010) with the ranking of a subset of 32 countries from the Luxembourg Income Study (LIS) [40], but there was no significant correlation (ρ = 0.114, p = 0.571) with the ranking of a subset of 27 countries from the European Union regarding single-parent families but not single-mother families [60].

Studies have focused on the rise of single motherhood in the EU [62] and the USA [61]. Over the study period, we observed an increase in single-mother families in 17 of 28 HBSC countries. However, regarding the data from all countries, the change between the surveys waves in 1994 and 2018 was not significant. Instead of this change, across all countries we observed a significant rise in the proportion of adolescents living in a family of the other structure type. In several countries, such as Greenland and England, this figure was over 20%. Although the World Family Map (2017) shows that the share of children living without parents is more common in African countries, there is also data on countries in the European region in previous years: for example, 5% in Italy (2014), 5% in Hungary (2001), and 4% in Romania (2004) [41]. So, it is clear that recently, the situation may have changed a lot. This phenomenon is related to the growing incidence of parental divorce and the subsequent establishment of new family types and living arrangements for young people. For instance, a child may be part of two families, when both biological parents establish new households after separation. A child may live part of the time with one biological parent and the remaining time with the other biological parent [1,49]. It is also the case that children are often raised by other relatives, either for their own good (e.g., fostered to an aunt who lives near a good school) or for the benefit of their host family (e.g., fostered in need of domestic labour) [42]. The increasing frequency of births outside of marriage, especially among young women, may also be one of the reasons why there has been an increase in the share of children who do not live with both biological parents. This may also be the case when the child is cared for by grandparents in order to help the parents to continue their studies and careers [41].

A specific reason for adolescents living without any of their parents could be that their parents go to work abroad due to difficult living conditions in their country. At the time of the survey (in 2014 and 2018), this phenomenon was more common in countries in Eastern Europe [63]. It corresponds to our findings, which showed that the number of adolescents living in families without parents has recently increased in Estonia, Latvia, Lithuania, and Russia. An optional question regarding the reasons why adolescents did not live with both parents also confirmed that a high proportion of adolescents declared that their parents work far away (it is unfortunate that this question was included in the HBSC survey in only four countries from Eastern Europe—Lithuania, Armenia, Moldova, and Ukraine).

Communication is a key modality by which the family functions as a protective health asset for adolescent development [64]. Difficult communication or a complete lack of communication due to the absence of the mother and father increases the probability of emotional and behavioural problems [65]. For many children whose parents work abroad, “home” means the absence of parents, emotional and psychological distress, and often physical and mental health problems [63,66]. Therefore, it deserves further and more detailed study.

Strengths and Limitations

One the strengths of the present analysis is that this study provided detailed analysis of the HBSC data on the composition of families in the year 2018 in 44 European countries and Canada and a trends analysis of the changes in family structures in 28 countries over the past three decades. Another strength is that the data was collected from the adolescent’s life perspective, making it more suitable for an assessment of adolescent health, well-being, and behaviour compared to official statistics. This study presents the adolescents’ perspectives and views on the types of families in which they live, and this has a strong added value to official statistics on family compositions and dynamics.

We also hereby acknowledge several limitations of the present study. First, the present study relied only on the adolescents’ self-reported data and these reports may have been subject to a potential response bias. The question on the composition of the family may have been sensitive for children who lived with one parent or with other relatives. For example, using sensitive questions, such as the question regarding with whom the adolescent lived with, can be affected by the possibility of a social fear bias in young people’s responses. The sensitivity was minimized by an effort to ensure the strict anonymity of respondents. Second, the classification of family structure types may have been a limitation, which was based on the adolescents’ responses. For example, some respondents indicated that they lived with their mother, father, and stepfather. According to our methodology, these responses were classified as an intact family, but it is also likely that the family couple was already in the process of separation. Third, the question regarding the reasons why the adolescent did not live with both parents was only asked in four countries participating in the HBSC study. After all, there was no clarification of the “other family structure”.

5. Conclusions

Over the last three decades, family structures in the European and North American regions have shown distinct patterns and changed considerably. The model of the intact family has been challenged by the recent trends of increasing family union instability and physical separation of families due to the emigration of parents or living apart.

The present study demonstrates how the adolescent’s perspective can contribute to better understanding the demographic trends and provide deeper knowledge on the changing sphere of family and parenthood. Its findings may have implications for cross-national adjustment of adolescent health, well-being, and behaviour by family structure, and for a critical analysis of socioeconomic family resources.