- Split View
-
Views
-
Cite
Cite
Agnete S. Dissing, Nadya Dich, Anne-Marie Nybo Andersen, Rikke Lund, Naja H Rod, Parental break-ups and stress: roles of age & family structure in 44 509 pre-adolescent children, European Journal of Public Health, Volume 27, Issue 5, October 2017, Pages 829–834, https://doi.org/10.1093/eurpub/ckx040
- Share Icon Share
Abstract
Parental break-up is wide spread, and the effects of parental break-up on children’s well-being are known. The evidence regarding child age at break-up and subsequent family arrangements is inconclusive. Aim: to estimate the effects of parental break-up on stress in pre-adolescent children with a specific focus on age at break-up and post-breakup family arrangements.
We used data from the Danish National Birth Cohort. Participants included 44 509 children followed from birth to age 11. Stress was self-reported by children at age 11, when the children also reported on parental break-up and post break-up family arrangements.
Twenty-one percent of the children had experienced a parental break-up at age 11, and those who had experienced parental break-up showed a higher risk of stress (OR:1.72, 95%CI:1.55;1.91) regardless of the child’s age at break-up. Children living in a new family with stepparents (OR = 1.63, 95%CI:1.38;1.92), or shared between the parents (OR = 1.48, 95%CI:1.26;1.75) reported higher stress than children of intact families. Single parent families reported markedly higher stress levels than children in intact families (OR = 2.18, 95%CI:1.90;2.50) and all other family types. Children who were satisfied with their living arrangements post-break-up reported the same stress level as children living in intact families (OR = 1.01, 95%CI:0.86;1.18).
Children who experience parental break-up have higher stress levels, also many years after the break-up, and those living in a single parent household post break-up seem to be most vulnerable. Living arrangements post-breakup should be further investigated as a potential protective factor.
Introduction
Approximately 1 million divorces are recorded every year in Europe1 and another 850 000 in the USA.2 It is estimated that slightly more than half of all divorces involve children under the age of 18.3 In addition, an increasing number of parents are not officially married,1 and if they break up, their children face similar life changes as children of married parents getting a divorce. Parental break-up is in most cases due to long-term conflicts in the family, and the detrimental effects of divorce and the underlying conflicts on the mental well-being of children are well documented.4,5 Considering that stress in childhood and early puberty is found to be a risk factor for later mental health disorders,6–8 the high number of children experiencing parental break-up constitute a particular public health concern. Hence, it is important to identify vulnerable children who may require extra support, and identify protective factors that may mitigate the negative effects of parental break-up.
Age at parental break-up might be an important factor when considering vulnerability to stress in children.9–17 The first years of life is a period characterized by rapid development of physical, social and mental abilities; a process where the family environment plays a crucial role.18 Both classical psychological theory as well as neuro-psychological research has emphasized caregiving and attachment in the first years of life as important for both physical and mental development in general,19 as well as for the development of stress specific systems such as regulations of cortisol levels.20,21 As the effect of divorce and parental conflicts on children’s mental well-being is suggested to be partly mediated by changes in attachment patterns,22 it is likely that the first years of life constitute a sensitive period, where family conflicts and parental break-up, may have a stronger impact on children’s later well-being than in other periods of the childhood. However, the current evidence regarding the role of child age at divorce is sparse, and many of the previous studies on child-age at divorce are characterized by using small or highly selected samples,9,10,16 very broad age groups (e.g. ages 0–15) or not covering the first years of life.11,13 Further, none of the identified studies accounted for maternal mental health even though this factor might play an important role for the relationship between parental break-up and child stress. Maternal mental health is likely to affect offspring well-being via both genetic pre-dispositions and environmental factors such as parenting style, as well as mental health problems are likely to also cause marital problems.23 Hence, it is not established whether children experiencing parental break-up in early life constitute a vulnerable group in need of extra support.
Post-divorce circumstances such as those related to the establishment of a new family structure may also be important for the well-being of the child,24–27 and it is suggested that children in single-parent families is a vulnerable group.28 An increasing number of children also live in a shared arrangement between their mother and father in a joint physical custody, which might also pose a stressful situation for the child as they are expected to adapt to two family environments.29 The child’s own involvement in and consent with the post-divorce living arrangements may also be important, but is seldom considered in practice settings or in research.24 Further exploration of living arrangements post break-up might provide more insight into such protective factors.
Using data from a large population-based cohort study with 44 509 children followed from birth, we aimed at investigating the effects of parental break-up on children’s stress in pre-adolescence. We extend the existing literature by specifically testing the hypothesis that children experiencing parental break-up in the first years of life exhibit higher levels of stress in pre-adolescence compared to children that experience break-up in other periods of childhood. We also aimed at identifying post break-up protective factors, such as family structure and the child’s satisfaction with the living arrangements. We will further, unlike previous studies, comprehensively adjust for maternal mental health.
Methods
Study population
We used data from the Danish National Birth Cohort (DNBC). The study design and response rates of DNBC have been described elsewhere.30,31 Women from all regions of Denmark were recruited at their first pregnancy visit to their general practitioner. In total, 92 274 women were recruited, which corresponds to approximately 60% of all invited women. When the child was 6 months old, the women were invited for a computer-controlled telephone interview. In the period from 2010 to 2014, the children of all recruited mothers were invited by letter to respond to an internet-based follow-up questionnaire (DNBC-11), where the majority of the children (82%) were 11 years of age. In total, 49 464 children of the 93 961 invited were followed up at the DNBC-11 corresponding to 52.6% of the original cohort. The sample for the present study included 44 509 children in the DNBC-11 who had full information on considered co-variables in the main analyses. To better understand selection mechanisms into the study population, we compared participants with children invited to the DNBC-11 using available information on parental co-habitation status registered in the Danish national registries in advance of the invitation to the DNBC-11. Parental co-habitation was associated with a higher response rate to the DNBC-11 (co-habiting: 51.5% vs. single: 42.7%, P < 0.001).
Measurements
Parental break-up
In DNBC-11, Parental break-up was evaluated by children reporting whether their parents were divorced or had stopped living together. Further, the children were asked to indicate how old they were when the parents had stopped living together. Age at parental break-up was categorized into three developmental stages: 0–3 (infancy and toddler), 4–8 years (childhood), and 9–11+ years (pre-adolescence). Family structure was evaluated by children reporting who they were living with most of the time. Further, children were asked to indicate if they were satisfied with living in the specified family structure (possible answers were: the arrangement suits me very well; the arrangement is all right; no, I wish it was different).
Child stress
Stress in children was assessed in DNBC-11 by the Stress in Children Questionnaire (SiC)-a 21-item validated instrument covering distress, emotional states, psychosomatic symptoms, school-related experiences and social support from peers. The SiC questionnaire was validated in a population of Swedish children in the ages 9–12 years, where it showed good reliability (Chronbach’s alfa = 0.86) and a high content validity.32 We removed one item referring to support from a grown up as we were concerned that this item was strongly correlated to the main exposure. In order to identify high stress in children, we dichotomized the SiC scale at a cut-off pre-defined by the authors of the SiC-questionnaire, and a score of 2.5 or higher on the scale ranging from one to four was considered as high child stress.
Co-variates
Information on maternal age was obtained at pregnancy. Age and sex of the child were obtained from the Danish Medical Birth Registry. Information on parental socio-economic position was collected during pregnancy and was assessed as the highest socio-economic position of either the father or the mother based on the occupational work function and the skills required to carry out this occupation. This information was grouped according to the Danish version of the International Standard Classification of Occupations.33,Maternal mental health history was assessed during pregnancy as lifetime psychiatric diagnosis using the following items: Have you ever had a psychiatric disorder or bad nerves? If so, the women were asked to indicate whether they had been diagnosed by a doctor and with which specific psychiatric diagnose. Maternal depressive symptoms were measured using items from the symptoms checklist-9234 at the time when the child was 6 months old.
Analytical strategy
A priori knowledge was used to graph causal assumptions of the investigated relationship. From these assumptions, we chose the efficient set of confounding variables using the rules for Directed Acyclic Graphs.35 We were aware that the investigated exposure variables were closely intertwined by nature, and to avoid co-linearity we constructed models for each investigated exposure variable: one crude model, one model adjusting for age and sex, maternal age, parental socio-economic position, and a third model adjusting additionally for maternal mental health history. Odds ratios (OR) and 95% confidence intervals (95% CI) for the associations between exposure variables and child stress were estimated using logistic regression. We conducted the following three sensitivity analyses: (i) Maternal depressive symptoms might constitute a better measure of maternal mental health than lifetime psychiatric diagnosis. However, the depressive symptoms variable had a considerable amount of missing information, and hence we chose to adjust for this variable in a separate analysis based on a sub-population with full information on this variable (N = 36 324). (ii) As attachment might be particularly important before the age of one, we conducted an analysis estimating the effect of parental break-up before the age of one on later child stress. (iii) Children of parents with broken partnerships were less likely to participate in the DNBC-11, and we were further worried that children with high stress were also less likely to participate as this could bias the results. Hence, we quantitatively assessed the potential impact this selection might have had on the results.36 (Supplementary table S1). All analyses were conducted in R version 3.0.2.
Ethical approval
The DNBC has been approved by the scientific ethics committee in Denmark and this specific study is carried out in keeping with regulations from the Danish Data Protection Agency.
Results
Population characteristics and child stress
Four percent of the children reported a high level of stress at age 11, and 21% of the children reported that their parents did not live together. Further distributions of the study characteristics are shown in table 1. The mean age of the children was 11.2 years and a few more girls (52%) than boys (48%) participated in DNBC-11. Most mothers gave birth between ages 23 and 34 years, and had medium to high socio-economic position. Approximately 7% of the mothers had a history of mental disorder. Broken partnership was most common for mothers who were in the youngest age group, unskilled parents and mothers with a mental disorder. Girls and children of young mothers, mothers with a mental disorder, and unskilled parents were in higher risk of reporting stress at age 11 (Supplementary table S2).
. | . | Parental broken partnershipb . | |
---|---|---|---|
. | N (Total)a . | No N (%) . | Yes N (%) . |
Sex | |||
Female | 23 145 (52.0) | 17 707 (78.6) | 4818 (21.4) |
Male | 21 364 (48.0) | 16 838 (80.5) | 4067 (19.5) |
Age | |||
10 years | 46 (0.1) | 38 (84.4) | 7 (15.6) |
11 years | 36 812 (82.7) | 28 730 (79.9) | 7212 (20.1) |
12 years | 6906 (15.5) | 5216 (77.6) | 1504 (22.4) |
13 years | 737 (1.7) | 558 (78.0) | 157 (22.0) |
14 years | 8 (0.0) | 3 (37.5) | 5 (62.5) |
Maternal age at pregnancy | |||
15–22 years | 1628 (3.7) | 886 (57.2) | 664 (42.8) |
23–28 years | 17 399 (39.1) | 13 378 (78.4) | 3684 (21.6) |
29–34 years | 19 971 (44.9) | 15 917 (81.5) | 3605 (18.5) |
35–40 years | 5256 (11.8) | 4167 (82.3) | 894 (17.7) |
41–46 years | 255 (0.6) | 197 (83.8) | 38 (16.2) |
Parental socio economic position | |||
University degree/work with leadership responsibility | 11 573 (26.0) | 9594 (84.2) | 1802 (15.8) |
Medium educated | 15 287 (34.3) | 12 224 (81.7) | 2731 (18.3) |
Skilled work | 11 215 (25.2) | 8426 (76.9) | 2536 (23.1) |
Unskilled work | 5086 (11.4) | 3407 (70.0) | 1462 (30.0) |
In education | 1046 (2.4) | 714 (73.3) | 260 (26.7) |
Outside labor market/not classified | 302 (0.7) | 180 (65.7) | 94 (34.3) |
Maternal mental health history | |||
No mental health disorder | 41440 (93.1) | 32481 (80.2) | 8019 (19.8) |
Self-reported disorder (no contact with health professional) | 249 (0.6) | 168 (70.6) | 70 (29.4) |
Anxiety/Affective disorder/suicide attempts/psychosis | 1511 (3.4) | 1008 (69.8) | 436 (30.2) |
Other disorders | 1309 (2.9) | 888 (71.2) | 360 (28.8) |
Maternal depressive symptomsc | |||
Perceive future as hopeless | |||
No | 32 647 (89.9) | 25 856 (81.0) | 6082 (19.0) |
Some | 3035 (8.4) | 2141 (73.1) | 788 (26.9) |
Much | 624 (1.7) | 388 (65.3) | 206 (34.7) |
Missing | 8203 | 6160 | 1809 |
Feeling sad or depressed | |||
No | 25 412 (70.0) | 20 382 (82.0) | 4479 (18.0) |
Some | 9665 (26.6) | 7202 (76.4) | 2226 (23.6) |
Much | 1229 (3.4) | 801 (68.3) | 371 (31.7) |
Missing | 8203 | 6160 | 1809 |
Feeling overloaded | |||
No | 31343 (86.3) | 24726 (80.7) | 5918 (19.3) |
Some | 4393 (12.1) | 3299 (77.2) | 973 (22.8) |
Much | 570 (1.6) | 363 (66.4) | 184 (33.6) |
Missing | 8203 | 6157 | 1810 |
Total | 44 509 (100) | 34 545 (79.6) | 8885 (20.4) |
. | . | Parental broken partnershipb . | |
---|---|---|---|
. | N (Total)a . | No N (%) . | Yes N (%) . |
Sex | |||
Female | 23 145 (52.0) | 17 707 (78.6) | 4818 (21.4) |
Male | 21 364 (48.0) | 16 838 (80.5) | 4067 (19.5) |
Age | |||
10 years | 46 (0.1) | 38 (84.4) | 7 (15.6) |
11 years | 36 812 (82.7) | 28 730 (79.9) | 7212 (20.1) |
12 years | 6906 (15.5) | 5216 (77.6) | 1504 (22.4) |
13 years | 737 (1.7) | 558 (78.0) | 157 (22.0) |
14 years | 8 (0.0) | 3 (37.5) | 5 (62.5) |
Maternal age at pregnancy | |||
15–22 years | 1628 (3.7) | 886 (57.2) | 664 (42.8) |
23–28 years | 17 399 (39.1) | 13 378 (78.4) | 3684 (21.6) |
29–34 years | 19 971 (44.9) | 15 917 (81.5) | 3605 (18.5) |
35–40 years | 5256 (11.8) | 4167 (82.3) | 894 (17.7) |
41–46 years | 255 (0.6) | 197 (83.8) | 38 (16.2) |
Parental socio economic position | |||
University degree/work with leadership responsibility | 11 573 (26.0) | 9594 (84.2) | 1802 (15.8) |
Medium educated | 15 287 (34.3) | 12 224 (81.7) | 2731 (18.3) |
Skilled work | 11 215 (25.2) | 8426 (76.9) | 2536 (23.1) |
Unskilled work | 5086 (11.4) | 3407 (70.0) | 1462 (30.0) |
In education | 1046 (2.4) | 714 (73.3) | 260 (26.7) |
Outside labor market/not classified | 302 (0.7) | 180 (65.7) | 94 (34.3) |
Maternal mental health history | |||
No mental health disorder | 41440 (93.1) | 32481 (80.2) | 8019 (19.8) |
Self-reported disorder (no contact with health professional) | 249 (0.6) | 168 (70.6) | 70 (29.4) |
Anxiety/Affective disorder/suicide attempts/psychosis | 1511 (3.4) | 1008 (69.8) | 436 (30.2) |
Other disorders | 1309 (2.9) | 888 (71.2) | 360 (28.8) |
Maternal depressive symptomsc | |||
Perceive future as hopeless | |||
No | 32 647 (89.9) | 25 856 (81.0) | 6082 (19.0) |
Some | 3035 (8.4) | 2141 (73.1) | 788 (26.9) |
Much | 624 (1.7) | 388 (65.3) | 206 (34.7) |
Missing | 8203 | 6160 | 1809 |
Feeling sad or depressed | |||
No | 25 412 (70.0) | 20 382 (82.0) | 4479 (18.0) |
Some | 9665 (26.6) | 7202 (76.4) | 2226 (23.6) |
Much | 1229 (3.4) | 801 (68.3) | 371 (31.7) |
Missing | 8203 | 6160 | 1809 |
Feeling overloaded | |||
No | 31343 (86.3) | 24726 (80.7) | 5918 (19.3) |
Some | 4393 (12.1) | 3299 (77.2) | 973 (22.8) |
Much | 570 (1.6) | 363 (66.4) | 184 (33.6) |
Missing | 8203 | 6157 | 1810 |
Total | 44 509 (100) | 34 545 (79.6) | 8885 (20.4) |
The study population was defined as having full response to the co-variables: age, sex, maternal age at pregnancy, parental socio economic position, and maternal mental health history.
373 children of parents that never lived together, 378 children of a deceased parent, 233 children living in foster home are excluded from this variable. 95 missing in this variable.
Reported when the child was 6 months old.
. | . | Parental broken partnershipb . | |
---|---|---|---|
. | N (Total)a . | No N (%) . | Yes N (%) . |
Sex | |||
Female | 23 145 (52.0) | 17 707 (78.6) | 4818 (21.4) |
Male | 21 364 (48.0) | 16 838 (80.5) | 4067 (19.5) |
Age | |||
10 years | 46 (0.1) | 38 (84.4) | 7 (15.6) |
11 years | 36 812 (82.7) | 28 730 (79.9) | 7212 (20.1) |
12 years | 6906 (15.5) | 5216 (77.6) | 1504 (22.4) |
13 years | 737 (1.7) | 558 (78.0) | 157 (22.0) |
14 years | 8 (0.0) | 3 (37.5) | 5 (62.5) |
Maternal age at pregnancy | |||
15–22 years | 1628 (3.7) | 886 (57.2) | 664 (42.8) |
23–28 years | 17 399 (39.1) | 13 378 (78.4) | 3684 (21.6) |
29–34 years | 19 971 (44.9) | 15 917 (81.5) | 3605 (18.5) |
35–40 years | 5256 (11.8) | 4167 (82.3) | 894 (17.7) |
41–46 years | 255 (0.6) | 197 (83.8) | 38 (16.2) |
Parental socio economic position | |||
University degree/work with leadership responsibility | 11 573 (26.0) | 9594 (84.2) | 1802 (15.8) |
Medium educated | 15 287 (34.3) | 12 224 (81.7) | 2731 (18.3) |
Skilled work | 11 215 (25.2) | 8426 (76.9) | 2536 (23.1) |
Unskilled work | 5086 (11.4) | 3407 (70.0) | 1462 (30.0) |
In education | 1046 (2.4) | 714 (73.3) | 260 (26.7) |
Outside labor market/not classified | 302 (0.7) | 180 (65.7) | 94 (34.3) |
Maternal mental health history | |||
No mental health disorder | 41440 (93.1) | 32481 (80.2) | 8019 (19.8) |
Self-reported disorder (no contact with health professional) | 249 (0.6) | 168 (70.6) | 70 (29.4) |
Anxiety/Affective disorder/suicide attempts/psychosis | 1511 (3.4) | 1008 (69.8) | 436 (30.2) |
Other disorders | 1309 (2.9) | 888 (71.2) | 360 (28.8) |
Maternal depressive symptomsc | |||
Perceive future as hopeless | |||
No | 32 647 (89.9) | 25 856 (81.0) | 6082 (19.0) |
Some | 3035 (8.4) | 2141 (73.1) | 788 (26.9) |
Much | 624 (1.7) | 388 (65.3) | 206 (34.7) |
Missing | 8203 | 6160 | 1809 |
Feeling sad or depressed | |||
No | 25 412 (70.0) | 20 382 (82.0) | 4479 (18.0) |
Some | 9665 (26.6) | 7202 (76.4) | 2226 (23.6) |
Much | 1229 (3.4) | 801 (68.3) | 371 (31.7) |
Missing | 8203 | 6160 | 1809 |
Feeling overloaded | |||
No | 31343 (86.3) | 24726 (80.7) | 5918 (19.3) |
Some | 4393 (12.1) | 3299 (77.2) | 973 (22.8) |
Much | 570 (1.6) | 363 (66.4) | 184 (33.6) |
Missing | 8203 | 6157 | 1810 |
Total | 44 509 (100) | 34 545 (79.6) | 8885 (20.4) |
. | . | Parental broken partnershipb . | |
---|---|---|---|
. | N (Total)a . | No N (%) . | Yes N (%) . |
Sex | |||
Female | 23 145 (52.0) | 17 707 (78.6) | 4818 (21.4) |
Male | 21 364 (48.0) | 16 838 (80.5) | 4067 (19.5) |
Age | |||
10 years | 46 (0.1) | 38 (84.4) | 7 (15.6) |
11 years | 36 812 (82.7) | 28 730 (79.9) | 7212 (20.1) |
12 years | 6906 (15.5) | 5216 (77.6) | 1504 (22.4) |
13 years | 737 (1.7) | 558 (78.0) | 157 (22.0) |
14 years | 8 (0.0) | 3 (37.5) | 5 (62.5) |
Maternal age at pregnancy | |||
15–22 years | 1628 (3.7) | 886 (57.2) | 664 (42.8) |
23–28 years | 17 399 (39.1) | 13 378 (78.4) | 3684 (21.6) |
29–34 years | 19 971 (44.9) | 15 917 (81.5) | 3605 (18.5) |
35–40 years | 5256 (11.8) | 4167 (82.3) | 894 (17.7) |
41–46 years | 255 (0.6) | 197 (83.8) | 38 (16.2) |
Parental socio economic position | |||
University degree/work with leadership responsibility | 11 573 (26.0) | 9594 (84.2) | 1802 (15.8) |
Medium educated | 15 287 (34.3) | 12 224 (81.7) | 2731 (18.3) |
Skilled work | 11 215 (25.2) | 8426 (76.9) | 2536 (23.1) |
Unskilled work | 5086 (11.4) | 3407 (70.0) | 1462 (30.0) |
In education | 1046 (2.4) | 714 (73.3) | 260 (26.7) |
Outside labor market/not classified | 302 (0.7) | 180 (65.7) | 94 (34.3) |
Maternal mental health history | |||
No mental health disorder | 41440 (93.1) | 32481 (80.2) | 8019 (19.8) |
Self-reported disorder (no contact with health professional) | 249 (0.6) | 168 (70.6) | 70 (29.4) |
Anxiety/Affective disorder/suicide attempts/psychosis | 1511 (3.4) | 1008 (69.8) | 436 (30.2) |
Other disorders | 1309 (2.9) | 888 (71.2) | 360 (28.8) |
Maternal depressive symptomsc | |||
Perceive future as hopeless | |||
No | 32 647 (89.9) | 25 856 (81.0) | 6082 (19.0) |
Some | 3035 (8.4) | 2141 (73.1) | 788 (26.9) |
Much | 624 (1.7) | 388 (65.3) | 206 (34.7) |
Missing | 8203 | 6160 | 1809 |
Feeling sad or depressed | |||
No | 25 412 (70.0) | 20 382 (82.0) | 4479 (18.0) |
Some | 9665 (26.6) | 7202 (76.4) | 2226 (23.6) |
Much | 1229 (3.4) | 801 (68.3) | 371 (31.7) |
Missing | 8203 | 6160 | 1809 |
Feeling overloaded | |||
No | 31343 (86.3) | 24726 (80.7) | 5918 (19.3) |
Some | 4393 (12.1) | 3299 (77.2) | 973 (22.8) |
Much | 570 (1.6) | 363 (66.4) | 184 (33.6) |
Missing | 8203 | 6157 | 1810 |
Total | 44 509 (100) | 34 545 (79.6) | 8885 (20.4) |
The study population was defined as having full response to the co-variables: age, sex, maternal age at pregnancy, parental socio economic position, and maternal mental health history.
373 children of parents that never lived together, 378 children of a deceased parent, 233 children living in foster home are excluded from this variable. 95 missing in this variable.
Reported when the child was 6 months old.
Parental break-up and the child’s age at break-up
Adjusting for potential confounding, children of all ages experiencing parental break-up had a higher risk of reporting stress in pre-adolescence (OR=1.72, 95%CI= 1.55;1.91) compared to children who live with both of their parents (table 2). Correcting this estimate (unadjusted) for assumed selection mechanisms suggested only some influence of the selection, and did not change the overall conclusion of the result (Supplementary table S1). It was apparent that children experiencing parental break-up earliest in life (age group 0–3 years) were not more vulnerable to stress compared to children experiencing parental break-up in later ages, which was contrary to our hypothesis (table 2). This was also true for children experiencing parental break-up before the age of one (Supplementary table S3). Adjusting for maternal depressive symptoms measured with the symptom-checklist in a subpopulation (N = 36 324) did not change the results considerably (Supplementary table S4).
. | N (total %) . | SiC > 2.5 N (%) . | Unadjusted . | Fully adjusted . | Fully adjusted + Maternal mental health . | |||
---|---|---|---|---|---|---|---|---|
. | . | . | OR . | [95%CI] . | OR . | [95%CI] . | OR . | [95%CI] . |
Parental broken partnershipa | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
Parents do not live together | 8885 (20.5) | 575 (6.5) | 1.88 | [1.69;2.08] | 1.75 | [1.58;1.94] | 1.72 | [1.55;1.91] |
Missing | 95 | |||||||
Age at parental break-upa | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 0.47 | [0.40;0.56] | 0.52 | [0.44;0.62] | 0.54 | [0.45;0.64] |
0–3 years | 2264 (5.2) | 165 (7.3) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
4–8 years | 4280 (9.9) | 260 (6.1) | 0.82 | [0.67;1.00] | 0.86 | [0.70;1.05] | 0.88 | [0.71;1.07] |
9–11+ years | 2341 (5.4) | 150 (6.4) | 0.87 | [0.69;1.09] | 0.93 | [0.74;1.17] | 0.95 | [0.75;1.19] |
Missing | 95 |
. | N (total %) . | SiC > 2.5 N (%) . | Unadjusted . | Fully adjusted . | Fully adjusted + Maternal mental health . | |||
---|---|---|---|---|---|---|---|---|
. | . | . | OR . | [95%CI] . | OR . | [95%CI] . | OR . | [95%CI] . |
Parental broken partnershipa | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
Parents do not live together | 8885 (20.5) | 575 (6.5) | 1.88 | [1.69;2.08] | 1.75 | [1.58;1.94] | 1.72 | [1.55;1.91] |
Missing | 95 | |||||||
Age at parental break-upa | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 0.47 | [0.40;0.56] | 0.52 | [0.44;0.62] | 0.54 | [0.45;0.64] |
0–3 years | 2264 (5.2) | 165 (7.3) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
4–8 years | 4280 (9.9) | 260 (6.1) | 0.82 | [0.67;1.00] | 0.86 | [0.70;1.05] | 0.88 | [0.71;1.07] |
9–11+ years | 2341 (5.4) | 150 (6.4) | 0.87 | [0.69;1.09] | 0.93 | [0.74;1.17] | 0.95 | [0.75;1.19] |
Missing | 95 |
Fully adjusted: Adjusted for maternal age at conception, sex of the child, age of child, parental socio economic position. OR = odds ratio, [95% CI] = 95% confidence intervals, [Ref] = reference group; SiC = stress in children scale.
Excluded from this variable: 373 children of parents that never lived together, 378 children of deceased parents, 233 children living in foster homes or in institution.
. | N (total %) . | SiC > 2.5 N (%) . | Unadjusted . | Fully adjusted . | Fully adjusted + Maternal mental health . | |||
---|---|---|---|---|---|---|---|---|
. | . | . | OR . | [95%CI] . | OR . | [95%CI] . | OR . | [95%CI] . |
Parental broken partnershipa | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
Parents do not live together | 8885 (20.5) | 575 (6.5) | 1.88 | [1.69;2.08] | 1.75 | [1.58;1.94] | 1.72 | [1.55;1.91] |
Missing | 95 | |||||||
Age at parental break-upa | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 0.47 | [0.40;0.56] | 0.52 | [0.44;0.62] | 0.54 | [0.45;0.64] |
0–3 years | 2264 (5.2) | 165 (7.3) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
4–8 years | 4280 (9.9) | 260 (6.1) | 0.82 | [0.67;1.00] | 0.86 | [0.70;1.05] | 0.88 | [0.71;1.07] |
9–11+ years | 2341 (5.4) | 150 (6.4) | 0.87 | [0.69;1.09] | 0.93 | [0.74;1.17] | 0.95 | [0.75;1.19] |
Missing | 95 |
. | N (total %) . | SiC > 2.5 N (%) . | Unadjusted . | Fully adjusted . | Fully adjusted + Maternal mental health . | |||
---|---|---|---|---|---|---|---|---|
. | . | . | OR . | [95%CI] . | OR . | [95%CI] . | OR . | [95%CI] . |
Parental broken partnershipa | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
Parents do not live together | 8885 (20.5) | 575 (6.5) | 1.88 | [1.69;2.08] | 1.75 | [1.58;1.94] | 1.72 | [1.55;1.91] |
Missing | 95 | |||||||
Age at parental break-upa | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 0.47 | [0.40;0.56] | 0.52 | [0.44;0.62] | 0.54 | [0.45;0.64] |
0–3 years | 2264 (5.2) | 165 (7.3) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
4–8 years | 4280 (9.9) | 260 (6.1) | 0.82 | [0.67;1.00] | 0.86 | [0.70;1.05] | 0.88 | [0.71;1.07] |
9–11+ years | 2341 (5.4) | 150 (6.4) | 0.87 | [0.69;1.09] | 0.93 | [0.74;1.17] | 0.95 | [0.75;1.19] |
Missing | 95 |
Fully adjusted: Adjusted for maternal age at conception, sex of the child, age of child, parental socio economic position. OR = odds ratio, [95% CI] = 95% confidence intervals, [Ref] = reference group; SiC = stress in children scale.
Excluded from this variable: 373 children of parents that never lived together, 378 children of deceased parents, 233 children living in foster homes or in institution.
Family structure and satisfaction with living arrangements
Children living in all other family structures than living with both parents had a higher risk of reporting stress at age 11 (table 3). Children living with only one parent had approximately twice as high odds of reporting stress (OR = 2.18, 95%CI=1.90;2.50) compared to children living with both parents. It should be noted that compared to children living in a single-parent family, children living shared between the parents (OR = 0.69, 95%CI = 0.56–0.84) or living with a step-parent (OR=0.77, 95%CI=0.62–0.94) reported markedly lower stress levels. This pattern persisted when adjusting for maternal depressive symptoms in a sub-population (Supplementary table S4). Nine percent of children not living with both of their parent reported to be unsatisfied with their living arrangements and those children had a markedly higher risk (OR = 5.44, 95%CI=4.44;6.67) of reporting stress compared to children who live with both their parents. It should be noted, that approximately half of the children of divorced parents were satisfied with their living situation, and that these children did not have a higher risk of reporting stress (OR = 1.01, 95%CI = 0.86;1.18) (table 2).
. | N (total %) . | SiC > 2.5 N (%) . | Unadjusted . | Fully adjusted . | Fully adjusted + Maternal mental health . | |||
---|---|---|---|---|---|---|---|---|
. | . | . | OR . | [95%CI] . | OR . | [95%CI] . | OR . | [95%CI] . |
Family structurea | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
Child lives shared between mother and father | 3222 (7.4) | 168 (5.2) | 1.49 | [1.26;1.76] | 1.50 | [1.27;1.77] | 1.48 | [1.26;1.75] |
Child lives alone with mother or father | 3032 (7.0) | 241 (7.9) | 2.34 | [2.03;2.70] | 2.23 | [1.95;2.55] | 2.18 | [1.90;2.50] |
Child lives with mother or father and a new partner | 2631 (6.1) | 166 (6.3) | 1.83 | [1.54;2.15] | 1.65 | [1.40;1.95] | 1.63 | [1.38;1.92] |
Missing | 95 | |||||||
Satisfaction with living arrangementsa | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
Yes | 4848 (11.2) | 181 (3.7) | 1.05 | [0.89; 1.23] | 1.02 | [0.88; 1.19] | 1.01 | [0.86; 1.18] |
OK | 3372 (7.8) | 278 (8.2) | 2.44 | [2.13; 2.79] | 2.32 | [2.03; 2.66] | 2.28 | [1.99; 2.61] |
No | 665 (1.5) | 116 (17.4) | 5.73 | [4.64; 7.03] | 5.58 | [4.55; 6.84] | 5.44 | [4.44; 6.67] |
Missing | 95 |
. | N (total %) . | SiC > 2.5 N (%) . | Unadjusted . | Fully adjusted . | Fully adjusted + Maternal mental health . | |||
---|---|---|---|---|---|---|---|---|
. | . | . | OR . | [95%CI] . | OR . | [95%CI] . | OR . | [95%CI] . |
Family structurea | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
Child lives shared between mother and father | 3222 (7.4) | 168 (5.2) | 1.49 | [1.26;1.76] | 1.50 | [1.27;1.77] | 1.48 | [1.26;1.75] |
Child lives alone with mother or father | 3032 (7.0) | 241 (7.9) | 2.34 | [2.03;2.70] | 2.23 | [1.95;2.55] | 2.18 | [1.90;2.50] |
Child lives with mother or father and a new partner | 2631 (6.1) | 166 (6.3) | 1.83 | [1.54;2.15] | 1.65 | [1.40;1.95] | 1.63 | [1.38;1.92] |
Missing | 95 | |||||||
Satisfaction with living arrangementsa | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
Yes | 4848 (11.2) | 181 (3.7) | 1.05 | [0.89; 1.23] | 1.02 | [0.88; 1.19] | 1.01 | [0.86; 1.18] |
OK | 3372 (7.8) | 278 (8.2) | 2.44 | [2.13; 2.79] | 2.32 | [2.03; 2.66] | 2.28 | [1.99; 2.61] |
No | 665 (1.5) | 116 (17.4) | 5.73 | [4.64; 7.03] | 5.58 | [4.55; 6.84] | 5.44 | [4.44; 6.67] |
Missing | 95 |
Fully adjusted: adjusted for maternal age at conception, sex of the child, age of child, parental socio economic position. OR = odds ratio; [95% CI] = 95% confidence intervals; [Ref] = reference group; SiC = stress in children scale.
Excluded from this variable: 373 children of parents that never lived together, 378 children of deceased parents, 233 children living in foster homes or in institution.
. | N (total %) . | SiC > 2.5 N (%) . | Unadjusted . | Fully adjusted . | Fully adjusted + Maternal mental health . | |||
---|---|---|---|---|---|---|---|---|
. | . | . | OR . | [95%CI] . | OR . | [95%CI] . | OR . | [95%CI] . |
Family structurea | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
Child lives shared between mother and father | 3222 (7.4) | 168 (5.2) | 1.49 | [1.26;1.76] | 1.50 | [1.27;1.77] | 1.48 | [1.26;1.75] |
Child lives alone with mother or father | 3032 (7.0) | 241 (7.9) | 2.34 | [2.03;2.70] | 2.23 | [1.95;2.55] | 2.18 | [1.90;2.50] |
Child lives with mother or father and a new partner | 2631 (6.1) | 166 (6.3) | 1.83 | [1.54;2.15] | 1.65 | [1.40;1.95] | 1.63 | [1.38;1.92] |
Missing | 95 | |||||||
Satisfaction with living arrangementsa | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
Yes | 4848 (11.2) | 181 (3.7) | 1.05 | [0.89; 1.23] | 1.02 | [0.88; 1.19] | 1.01 | [0.86; 1.18] |
OK | 3372 (7.8) | 278 (8.2) | 2.44 | [2.13; 2.79] | 2.32 | [2.03; 2.66] | 2.28 | [1.99; 2.61] |
No | 665 (1.5) | 116 (17.4) | 5.73 | [4.64; 7.03] | 5.58 | [4.55; 6.84] | 5.44 | [4.44; 6.67] |
Missing | 95 |
. | N (total %) . | SiC > 2.5 N (%) . | Unadjusted . | Fully adjusted . | Fully adjusted + Maternal mental health . | |||
---|---|---|---|---|---|---|---|---|
. | . | . | OR . | [95%CI] . | OR . | [95%CI] . | OR . | [95%CI] . |
Family structurea | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
Child lives shared between mother and father | 3222 (7.4) | 168 (5.2) | 1.49 | [1.26;1.76] | 1.50 | [1.27;1.77] | 1.48 | [1.26;1.75] |
Child lives alone with mother or father | 3032 (7.0) | 241 (7.9) | 2.34 | [2.03;2.70] | 2.23 | [1.95;2.55] | 2.18 | [1.90;2.50] |
Child lives with mother or father and a new partner | 2631 (6.1) | 166 (6.3) | 1.83 | [1.54;2.15] | 1.65 | [1.40;1.95] | 1.63 | [1.38;1.92] |
Missing | 95 | |||||||
Satisfaction with living arrangementsa | ||||||||
Child lives with both parents | 34 545 (79.5) | 1228 (3.6) | 1 [Ref] | 1 [Ref] | 1 [Ref] | |||
Yes | 4848 (11.2) | 181 (3.7) | 1.05 | [0.89; 1.23] | 1.02 | [0.88; 1.19] | 1.01 | [0.86; 1.18] |
OK | 3372 (7.8) | 278 (8.2) | 2.44 | [2.13; 2.79] | 2.32 | [2.03; 2.66] | 2.28 | [1.99; 2.61] |
No | 665 (1.5) | 116 (17.4) | 5.73 | [4.64; 7.03] | 5.58 | [4.55; 6.84] | 5.44 | [4.44; 6.67] |
Missing | 95 |
Fully adjusted: adjusted for maternal age at conception, sex of the child, age of child, parental socio economic position. OR = odds ratio; [95% CI] = 95% confidence intervals; [Ref] = reference group; SiC = stress in children scale.
Excluded from this variable: 373 children of parents that never lived together, 378 children of deceased parents, 233 children living in foster homes or in institution.
Discussion
In this large cohort study following 44 509 children from birth into pre-adolescence, we found long-term associations between parental break-up and stress. The hypothesis that children younger than three are particularly vulnerable to parental break-up was not supported in the present study. Family structure and living arrangements after a break-up appear to be important for children’s well-being.
Parental break-up in early and later childhood
We find that parental break-ups are detrimental to the well-being of the involved children; a finding which is also well supported in the literature.4,5 The effects of parental break-up and the preceding conflicts are highly intertwined and a break-up will in many cases be an indicator of underlying stressful circumstances. Contrary to our expectations, children experiencing parental break-up in early life did not appear to be more vulnerable than children experiencing parental break-up in later childhood. Previous studies show mixed findings, and have identified both early and later childhood as sensitive periods for parental break-up,12,14–17 which may be attributed to differences in methodology and study populations. A reason for not detecting young age as a sensitive period in the current population could be that children experiencing the break-up in older ages might have been more affected by a recent parental break-up and a longer exposure to parental conflict. Further, the mechanisms increasing the vulnerability of children going through a parental break-up might also operate differently at different ages. Whereas reduced quality of parent-child relation following parental break-up might be a key mechanism in very young children,19 the child’s appraisal of the conflicts may play a more prominent role in older children with more advanced cognitive skills.37,38 We were not able to further explore age-related differences in the mechanisms with the data available and further research is needed to fully understand these mechanisms.
Family structure and satisfaction with living arrangements
We found living arrangements after parental break-up to be associated with stress in pre-adolescence. Children living with a single parent appeared to be a particularly vulnerable group, which is in line with findings from other studies.25,27 In addition to reduced economic resources available,39 children in this family type often suffer from a reduced contact to the non-custodial parent.24 Children in joint physical custody showed a lower stress level than children living with a single parent and one explanation may be that the child is more likely to maintain a close attachment to both parents in spite of the break-up.25 Further, it is suggested that parents with joint physical custody of the child are better at cooperating and hence experience less conflicts, which is beneficial to the child.24,29 Lower conflict levels and cooperation between parents post-divorce may also be an indicator of lower conflict levels pre-divorce, which might also explain why this group of children appears to be more resilient than children in other non-intact families. Both joint physical custody as well as single-parent families are arrangements that are likely to follow a recent break-up which might lead to a higher stress level for both of these groups. In spite of this, we note that the risk of stress is markedly higher for children living in single-parent families.
Approximately half of the children of parents not living together were satisfied with their living arrangements post break-up and these children did not have a higher stress level than children living in intact families. However, both satisfaction with living arrangements and stress were self-reported by the children, and it is therefore not possible to tell whether satisfaction with living arrangement reduces stress, or if both low stress level and satisfaction are indicators of a resilient group of children who cope better with the break-up of their parents. Nevertheless, the important implication of this finding is that higher levels of stress are not a universal consequence of parental break-up and a considerable proportion of children show good adjustment.
Strengths and limitations
We adjusted our analyses for maternal mental health; an important confounder, which have rarely been adjusted for in similar studies. However, as in any observational study we cannot exclude the possibility that other unmeasured factors such as e.g. paternal mental health are contributing to the observed associations.
We used a validated tool to measure child stress, which incorporates several aspects of stress reported by the child, giving us a good indicator of child mental well-being. Experiencing parental break-up is a life changing event that has such major implications that it seems likely that the child will be fairly capable at remembering the timing of parental break-up. Further, we considered the family structure from the perspective of the child, and as a consequence analysed co-habiting non-married and married parents in one group. Co-habiting childbearing does not appear to be socially patterned in the Nordic countries,40 and hence it is unlikely that this grouping had a large impact on the results.
Selection bias was a concern in this study, and we hypothesized that the most likely scenario was a lower likelihood of participation among children with parents not co-habiting and children with stress. We conducted a quantitative bias analysis showing that under this scenario, the observed estimate is likely to be an underestimation of the true effects, and that the result is fairly robust even to extreme versions of this scenario. Despite the large sample of the DNCB, the recruitment of the mothers into DNBC was to some extent dependent on socio-economic factors,30 which may compromise the generalizability of the results to families with severe social problems.
Conclusion
Children who experience parental break-up at any age are at higher risk of reporting stress in pre-adolescence compared to children living in intact families. Children living in a single parent household seem to be particularly vulnerable to stress. There appears to be a complex interplay between family relationships and family structure where stressful effects of parental break-up might be moderated by cooperative parents and the child having a sense of satisfaction with their living arrangement.
Supplementary data
Supplementary data are available at EURPUB online.
Funding
The Danish National Research Foundation established and funded the Danish Epidemiology Science Centre that initiated the Danish National Birth Cohort. Additional financial support for the DNBC was obtained from the Pharmacy Foundation, the Egmont Foundation, the March of Dimes Birth Defects Foundation, the Augustinus Foundation, and the Health Foundation. The 11-year follow-up was funded by the Danish Medical Research Council, the Lundbeck Foundation and a strategic grant from Department of Public Health, University of Copenhagen.
Conflicts of interest: None declared.
Parental break-up is widespread, and knowledge about factors that either increase or decrease the vulnerability of children is important for policies targeting children experiencing parental break-up.
Regardless of age at break-up, children of parental break-up report higher stress than children in intact families.
Children in single-parent families have higher stress than children in other family structures.
Children’s satisfaction with living arrangements might moderate the stressful effects of parental break-up.
Comments