The lifelong socioeconomic disadvantage of single-mother background - the Helsinki Birth Cohort study 1934–1944

Growing up with one parent has become increasingly common for children in the Western world. From 1980s to 2008 the share of single parent families in Europe rose from 10 to 21 % mainly due cohabitations and divorces but there are demographic differences in single parenthood between European countries [1]. However, in Europe and elsewhere low socioeconomic position and poverty is more common in single parent households, especially those headed by a woman [2, 3].

The lifelong consequences of low childhood socioeconomic position (SEP) has been extensively studied [4, 5]. In most studies parents’ low occupational status has been used as an indicator of childhood SEP [6, 7], while markers indicating other aspects of potential early life deprivation, such as mother’s marital status, have been less frequently used. However, numerous studies show that single parent family background is highly relevant determinant of health at least among children and young adults, among whom it is associated with adverse physical and mental health outcomes [8‐12] as well as poorer educational performance and idleness (being neither in school nor employed) [11‐14]. Less is known about the effects of single parent background across the life course, because most of the previous studies are performed in relatively young cohorts. Those few studies that have focused on lifelong effects of single motherhood have indicated that at least men born outside of marriage have poorer health and higher mortality and they are more likely to stay unmarried or become divorced [15‐18]. Some studies also suggest that people born out of wedlock attain lower adult socioeconomic position (SEP) than those born to a married mother [16, 17]. However, it is not clear how the lifelong consequences of being born to an unmarried mother are mediated through the other childhood socioeconomic factors, and results in these mediating associations are contradictory [10, 12, 19‐23].

Further, in many studies single parent families are also seen as a homogenous group, although it includes children who grow up with a single parent from birth to adult life, and children born to an unmarried mother who later marries and the child grows up with two parents or caregivers. Less attention has been paid to outcomes among different subgroups such as children born out of wedlock. Thus, a recent review concluded that research has not distinguished out of wedlock birth from single parenthood in general and called studies focusing on outcomes in different subgroups [13].

The main objective of this study was to examine whether being born out of wedlock in a society where marriage was the norm is associated with socioeconomic position and marital status in later life. The secondary objectives were to assess whether these associations are mediated through perinatal and childhood factors associated with being born to an unmarried mother and whether they are specific to children with no indication of a male caregiver rather than those who later had a male caregiver. In measuring socioeconomic position in adulthood we used different indicators of SEP from several time points across adult life.

This study is a part of the Helsinki Birth Cohort Study (HBCS), which includes 13345 people born in either of the two public maternity hospitals from 1934 to 1944. The HBCS has been described in detail [24] and has been approved by the Ethics Committee of the National Public Health Institute. Register data were linked with permissions from Statistics Finland and the Finnish Ministry of Social Affairs and Health. Early life data came from birth, child welfare clinic and school healthcare records.

Maternal and neonatal characteristics are shown in Table 1. Unmarried mothers were on average younger, shorter and had a lower body mass index in late pregnancy. They also had a lower occupational status than married women. Their offspring were on average lighter and shorter than children born to married mothers, they were born at an earlier gestational age and were more often firstborn. In families of unmarried mothers who later had a male caregiver, the male caregiver had on average a lower occupational status than male caregivers in families of married mothers. This difference was less strong than the difference in occupational status between unmarried and married mothers.

Table 3 presents odds ratios (ORs) and 95 % confidence intervals (CIs) from the multinomial logistic regression predicting the highest educational attainment, occupational status and maximum income in adulthood according to mother’s marital status at the time of birth. Compared to children of married mothers, children of unmarried mothers were more likely to have lower educational and occupational levels and income as adults. The associations were graded, so that the ORs increased with decreasing educational, occupational and income categories. They were little changed after adjusting for year of birth, sex, birth order, mother’s age, length of gestation and wartime evacuation status. Further adjustments with mother’s height and body mass index in late pregnancy attenuated the associations slightly (data not shown). In addition, after adjusting for mother’s occupational status the associations were somewhat attenuated, although most remained statistically significant. All associations were similar in men and women (p-values for interaction sex * unmarried mother >0.05).

We then compared different types of families based on mother’s marital status at birth and later presence of a male caregiver and his occupational status. As a reference group we used children who were born to married mothers and whose fathers/male caregivers were manual workers. Compared with these children, children born to an unmarried mother who had no male caregiver during childhood had the highest chance of ending up in the lower SEP groups in adulthood (Tables 4, 5 and 6). Also children born to unmarried mothers who later had a manual worker male caregiver were more likely to attain a lower adult SEP than corresponding children born to married mothers. The tables show that children who had an unmarried mother and who later had a male caregiver in a middle class profession had similar SEP prospects as the reference group of children of married mothers and manual worker fathers. However, these children attained a lower SEP than children of married mothers whose fathers had a middle class profession. After adjusting for mother’s occupational status the ORs attenuated a little). In other words, having a mother in a middle class profession, regardless of whether the child was born to a married or unmarried mother, seemed to increase the possibility of reaching a higher SEP in adulthood.

Children born to unmarried mothers were more likely to stay unmarried in adulthood compared with children of married mothers (Table 7). Again, the group with the highest odds comprised children born to an unmarried mother with no male caregiver during childhood. In analyses related to marital status in adulthood, we found no difference between any other family groups (Table 8). Children of unmarried mothers who ended up married in adulthood had a similar chance of divorce than those of married mothers.

We examined how mother’s marital status when giving birth was associated with offspring’s lifetime SEP and marital status. Our results showed that children born out of wedlock almost 80 years ago attained a lower SEP in adulthood and were more likely to remain unmarried than children of married mothers. The most disadvantaged group consisted of children born to an unmarried mother who did not have a male caregiver during childhood years. They were most likely to end up in lower SEP and to stay unmarried. Also children who were born to an unmarried mother and who later had a male caregiver attained a lower SEP than those born to married mothers. Furthermore, having a mother with a middle class profession increased the offspring’s probability of reaching a higher SEP in adulthood independently of mother’s marital status at birth.

Previous studies have reported an association between single parent background and lower SEP in later life [12, 14, 16, 17] but there are also studies showing no association after childhood socioeconomic circumstances are accounted for [21, 23]. However, the majority of this previous research has focused on outcomes in early adulthood [12, 22, 32, 33] and treated divorced, separated, never married as well as widowed, as if they were a homogeneous single parent group [21, 33‐35]. Fewer studies have focused only on children born out of wedlock and measured changes in SEP in later life [16, 17]. This kind of a study design has been used in a small number of studies from Sweden [16] and Denmark [17]. These studies have reported associations between out of wedlock birth and later life outcomes, including SEP as well as health, largely independent of childhood socioeconomic circumstances. Our results are in line with these studies. We found that out of wedlock birth was associated with worse SEP in adulthood according to all indicators. The associations we found, were not explained through socioeconomic circumstances in childhood even though mother’s occupation at birth, as a marker of childhood socioeconomic position, was positively related to the likelihood that a child would attain a higher SEP in adulthood. Previous Scandinavian studies have also reported that men, but not women, born out of wedlock tend to more often become widowers, experience divorce and stay unmarried themselves [15‐17]. In the present study this association between out of wedlock birth and the tendency to stay unmarried was found among both sexes. However, we found no association between out of wedlock birth and experiencing of divorce.

Our study adds to this previous literature by showing that the long-term effects of single parent background remained, in both sexes, even when adulthood SEP was measured several times during adult life and was observed by using three different indicators of SEP.

Several mechanisms could explain these associations. First, being born to an unmarried mother may have meant a lack of material resources, for example food, during the foetal period and throughout childhood. Intrauterine growth, for which size at birth in relation to gestational age serves as a marker, has an important role for the development of vital organs such as the brain and may have long-term consequences for later development [36, 37]. Second, the social stigma experienced by children born out of wedlock may have lead to stress [11, 13] and have an impact on their upbringing, self-identity and which may subsequently translate to later life consequences including SEP. From an evolutionary perspective, investments made for a child can differ greatly between single parent and two parent families. In two parent families both parents usually have more resources, e.g. income, time and attention, while a single parent may have to focus much of the social and material resources on earning a livelihood and possibly combating social stigma [11, 12, 38, 39]. If a single mother later is married it may improve children’s standard of living although relationships in stepfamilies may be tense [11, 40].

The strong associations between mother’s marital status and offspring SEP in adulthood we found are likely to contribute to the association between low childhood SEP and health in later life. However, our data also show that many people born to an unmarried mother gain a good education and occupation despite early disadvantage. Moreover, it is important to highlight that in any study out of wedlock birth should be understood as an indicator of social disadvantage in the specific historical context of the study period. Our cohort members were born between 1934 and 1944 during an era when marriage was the norm. Therefore no direct analogy can be drawn to birth outside marriage and single parenting in contemporary societies. There are, however, still many contemporary contexts, where single mothers may have limited material and social resources similarly to participants of the present study.

Our results deserve an additional note from a technical perspective. Father’s occupational status is a widely used indicator of childhood SEP in epidemiological studies. It is important to realise that a missing father’s occupation may indicate a child born to an unmarried mother. In such cases it may be wiser to include such children as a separate category rather than to exclude them because of missing data, assess SES by the single parent’s occupation, or to use a purely statistical imputation technique. The limitations of the HBCS have been discussed [41]. Data on marital status, SEP and birth characteristics came from reliable records. While SEP in childhood was based only on occupational information, we have used both the mother’s and possible male caregiver’s occupations and classified occupational status based on modern standards of classifying historical occupational data. We have no data, however, on the remaining family members other than their number. Moreover, the adult occupational data are available from year 1970 onwards, when the cohort members were aged 26–36 years, precluding the assessment of socio-economic trajectories in adult life. We have no data on cohabitation without marriage. This is unlikely to be a major limitation for our main exposure, as marriage was the norm when the study participants were born, but constitutes a limitation for marital status as an outcome. Our study population may not be representative of all people living in Helsinki at that time. However, the distribution of occupational categories is consistent with the general occupational distribution in Helsinki 80 years ago among married HBCS families [42]. Child welfare clinic services were voluntary and free of charge and our data indicate that these clinics were attended by children from all socioeconomic groups. It is also possible that some unmarried mothers were recorded as married in the records and thus the married group may include a proportion of “false negatives”. This would be expected, if anything, to reduce the differences between children of married and unmarried mothers.

This life course study shows that children born out of wedlock carry a socioeconomic disadvantage throughout life. As compared with children born to married mothers, they have approximately three-fold odds of ending up in the lowest than in the highest educational and occupational categories. Most likely to end up in these categories are children born to unmarried mothers who have no male caregiver during childhood. These associations are not explained by other socioeconomic factors as indicated by mother’s and possible male caregiver’s occupational statuses. This disadvantage starting in early life is likely to have a substantial effect on lifetime health.