Teenage parenthood among child welfare clients: A Swedish national cohort study of prevalence and odds

ARTICLE IN PRESS

Journal of Adolescence Journal of Adolescence 30 (2007) 97–116 www.elsevier.com/locate/jado

Teenage parenthood among child welfare clients: A Swedish national cohort study of prevalence and odds Bo Vinnerljunga,b,, Eva Franze´na, Maria Danielssona b

a Center for Epidemiology, National Board of Health and Welfare, S 106 30 Stockholm, Sweden Institute for Evidence-Based Social Work Practice, National Board of Health and Welfare, S 106 30 Stockholm, Sweden

Abstract To assess prevalence and odds for teenage parenthood among former child welfare clients, we used national register data for all children born in Sweden 1972–1983 (n ¼ 1; 178; 207), including 49,582 former child welfare clients with varying intervention experiences. Logistic regression models, adjusted for demographic, socio-economic and familial background factors, were used to estimate odds ratios. Among youth who received interventions in adolescence, 16–19% of the girls and 5–6% of the boys became teenage parents, compared to 3% for girls and 0.7% for boys without child welfare experiences. Youths who entered child welfare services in their teens had four- to fivefold adjusted odds for becoming a teenage parent. For other child welfare clients, adjusted odds were mostly twofold. Youth of both sexes who receive child welfare services in adolescence are a high-risk group for teenage parenthood. Child welfare agencies should, as a minimum, provide each individual client youth with access to birth control counselling and contraceptives. r 2005 The Association for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved. Keywords: Adolescent; Parenthood; Motherhood; Fatherhood; Child bearing; Child welfare; Foster care; Child care; Cohort study

Corresponding author. Tel.: +46 8 5555 3213; fax: +46 8 5555 3327.

E-mail addresses: [email protected] (B. Vinnerljung), [email protected] (E. Franze´n), [email protected] (M. Danielsson). 0140-1971/$30.00 r 2005 The Association for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.adolescence.2005.12.002

ARTICLE IN PRESS B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

98

Aims In this study, based on 12 national birth cohorts (n ¼ 1; 178; 207) born 1972–1983, we assess the prevalence of teenage parenthood among nearly 50,000 former child welfare clients of both sexes with varying experiences of child welfare interventions. We also examine the odds of child welfare clients becoming a parent during teenage years, compared to peers with similar socio-economic and demographic backgrounds. In the analysis, we make special efforts to describe and analyse outcomes for different subgroups within the heterogeneous child welfare client population.

Background Why it matters International medical and social research has consistently found that teenage parenthood is associated with higher risks for a number of negative outcomes for both mothers and children (especially for the youngest mothers and their children), usually reinforced when teenage parenthood is combined with a background of familial poverty (e.g. Olausson, Cnattingius, & Goldenberg, 1997):

 Adverse

 

 



pregnancy outcomes, like low birth weight, pre-term delivery, neonatal and postneonatal mortality (Botting, Rosato, & Wood, 1998; Fraser, Brockert, & Ward, 1995; Irvine, Bradley, Cupples, & Boohan, 1997; Olausson, Cnattingius, & Haglund, 1999). Teenage mothers also tend to seek antenatal health care at a later point in their pregnancy than nonadolescent women (Kaufman, 1996). Poorer mental health later in adulthood for mothers (e.g. Maskey, 1991) and lower life satisfaction (Hobcraft & Kiernan, 2001). Lost education opportunities for mothers, low socio-economic status (SES) later in adulthood, welfare dependency, unstable family situation over long periods of the life course and even increased risk of premature death, including death from suicide and violence (Chilman, 1980; Furstenberg, Brooks-Gunn, & Chase-Lansdale, 1989; Hardy, Astone, BrooksGunn, Shapiro, & Miller, 1998; Hobcraft & Kiernan, 2001; Kiernan, 1997; Manlove, 1997; Olausson, Haglund, Weitoft, & Cnattingius, 2001, 2004; Williams, Forbes, McIlwaine, & Rosenberg, 1987). Child maltreatment (e.g. Dixon, Browne, & Hamilton-Giachritsis, 2005; Lee & Goerge, 1999) and injuries among the children due to accidents (e.g. Hobcraft & Kiernan, 2001). Low educational attainments among offspring (Furstenberg, Levine, & Brooks-Gunn, 1990; Manlove, 1997), behavioural problems and higher rates of delinquency among offspring (review in Coley & Chase-Lansdale, 1998) and also increased risk for suicidal behaviour among offspring in adolescence and young adulthood (Lewinsohn, Rohde, & Seeley, 1994; Mittendorfer-Rutz, Rasmussen, & Wasserman, 2004). Intergenerational transmission of early parenthood, poverty and poor living conditions to offspring (e.g. Hardy et al., 1998; Kiernan, 1997; Serbin & Karp, 2004).

ARTICLE IN PRESS B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

99

Children and youth in care For young women, risk of teenage parenthood has been linked to a number of background factors. Examples are listed below. Most of these traits are also prominently over-represented in child welfare populations:

 Low     

SES or poverty in birth family (e.g. Kaufman, 1996; Manlove, 1997; Vikat, Rimpela¨, Kosunen, & Rimpela¨, 2002), a hallmark of children who receive in-home or out-of-home care (e.g. Bebbington & Miles, 1989; Vinnerljung, Hjern, & Lindblad, in press). Familial background of teenage parenthood (e.g. Kiernan, 1997; Manlove, 1997), common among children/youth who become child welfare clients in Sweden and elsewhere (e.g. Vinnerljung, 1996a; Vinnerljung et al., in press). Low educational attainments among the adolescents themselves (e.g. Kaufman, 1996; Serbin et al., 1998), a dismal characteristic of youth in out-of-home care (e.g. Vinnerljung, O¨man, & Gunnarson, 2005). History of childhood aggression and of antisocial behaviour in adolescence (Rutter, Giller, & Hagell, 1998; Serbin & Karp, 2004; Serbin et al., 1998), common traits among adolescent girls who receive child welfare interventions (e.g. Vinnerljung, Sallna¨s, & Kyhle Westermark, 2001). Absence of warm and involved parenting in birth home (e.g. Scaramella, Conger, Simons, & Whitbeck, 1998). History of being a victim of sexual abuse or sexual force (Coley & Chase-Lansdale, 1998; Polit, Morton, & White, 1989), although contradicting results have also been reported (e.g. Widom & Kuhns, 1996).

Several studies have shown that more adolescents from reasonably affluent circumstances tend to choose abortion after a pregnancy, compared to girls from disadvantaged families (e.g. Smith, 1993; Turner, 2004). In view of the social selection among child welfare clients, it is logical that this seems to be valid also for young females in out-of-home care (Corlyon & McGuire, 1999). Child welfare clients—irregardless of sex—also tend to be sexually active at an earlier age than their general population peers, and be less likely to use contraceptives effectively (Corlyon & McGuire, 1999; Polit et al., 1989). Corlyon and McGuire (1999) reported that both girls and boys in British out-of-home care were under increased peer pressure to engage in sexual relations, compared to a sample of same age boys and girls from similar socio-economic backgrounds, but living at home. Research findings on teenage fathers’ backgrounds is mostly consistent with what we know of adolescent boys who become child welfare clients (e.g. Fagot, Pears, Capaldi, Crosby, & Leve, 1998; Vinnerljung et al., 2001). Teenage fathers tend to come from socio-economic poor homes, often with a history of welfare dependency and familial teenage parenting (Coley & ChaseLansdale, 1998; Kiernan, 1997). Low educational attainments and school problems are very common (Bunting & McAuley, 2004; Furstenberg et al., 1989; Kiernan, 1997). Many have conduct disorders and are engaged in delinquency and substance abuse (Fagot et al., 1998; Kaufman, 1996; Miller & Moore, 1990). Other forms of risk-taking behaviour, including habitual unprotected sex, are common (e.g. Holmberg & Berg-Kelly, 2002).

ARTICLE IN PRESS 100

B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

Looking at the brief summary above, it is not surprising that studies of out-of-home care populations in adolescence or later in adult life have repeatedly found high rates of teenage childbearing. This has been the case in the UK (Biehal, Clayden, Stein, & Wade, 1992, 1995; Buchanan, 1999; Garnett, 1992; Wellings, Wadsworth, Johnson, Field, & MacDowall, 1999), the US (e.g. Barth, 1990; Cook, 1994; Courtney & Piliavin, 1998; McMillen & Tucker, 1999; Polit, Morton & White, 1989); Australia (Cashmore & Paxman, 1996; Mendes & Moslehuddin, 2004), Denmark (Christoffersen, 1993, 2005) and Sweden (e.g. Vinnerljung, 1996a; Vinnerljung, Sundell, Lo¨fholm, & Humlesjo¨, 2006). But most of these studies have been performed on small and local samples, thereby lacking the capacity to produce nationally representative results. Also, some have used out-of-home care samples dominated by youth from long-term foster care, a selected group among child welfare clients. Few have adjusted results for birth home related socio-economic and demographic background factors, and have subsequently not been able to estimate risks or odds compared to general population peers with similar socio-demographic backgrounds. Finally, none of these studies has, as far as we know, discriminated between substantially different subgroups within the heterogeneous child welfare population, for example, between those entering the system in young age and youth receiving services in adolescence. Using a sample of 12 national cohorts including almost 50,000 child welfare clients, and being able to adjust results for parental background of the adolescents, we try to address these shortcomings. National rates of teenage births Although research on teenage parents from different developed countries have yielded similar results on risks and outcomes, rates differ substantially between nations (e.g. Danielsson, Rogala, & Sundstro¨m, 2001; Jones et al., 1985). Sweden has one of the lowest teenage birth rates among the OECD countries (see Table 1), in spite of liberal attitudes to sex during adolescence (Darroch, Frost, & Singh, 2001; Jones et al., 1985). In comparison, rates are eight times higher for US peers and five times higher in the UK. Table 1 Teenage birth rates 1998 for 10 selected OECD countries Country

Teenage birth rate

USA UK Canada Australia Germany France Italy Sweden Netherlands Japan

52.1 30.8 20.2 18.4 13.1 9.3 6.6 6.5 6.2 4.6

Births per 1000 women 15–19 (Source: OECD, 2002).

ARTICLE IN PRESS B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

101

A comparative study from the late 1990s of adolescent sexuality and pregnancies in Sweden, France, Canada, the UK and the US found great differences in teenage pregnancy rates between nations (Danilesson, Rogala & Sundstro¨m, 2003; Darroch et al., 2001; cp. Santow & Bracher, 1999). But there were only minor variations in teenage sexual behaviour. What differed was knowledge about and access to birth control, and peer values about early versus postponed parenthood. A brief note on Swedish child welfare Looking at the relatively large quantitative scope of Swedish child welfare, it may be questionable to label it ‘‘a marginal phenomenon’’. In the capital of Stockholm, around 10% of all children born 1968–1975 received child welfare interventions of some kind during their formative years (Sundell, Vinnerljung, Lo¨fholm, & Humlesjo¨, 2004). Six to eight per cent for the entire country is a probable estimate (national data on all forms of interventions do not exist). In contrast to, for example, the UK, juvenile delinquency (up to age 18) is dealt with almost entirely within the jurisdiction of child welfare (see Heesle & Vinnerljung, 1999).

Method Sweden has a long tradition of national registers, covering the entire population, with highquality data for socio-economic indicators and for child welfare interventions. These registers are based on the individually unique 10-digit personal identification number that follows every Swedish resident from birth—or date of immigration—to death. They allow members of the same family to be linked over generations. Our study utilizes some of these registers, maintained by Statistics Sweden and by the National Board of Health and Welfare. Study population The study population consists of the entire Swedish born population born 1972–1983, still alive and residing in Sweden at age 20, identified in the Swedish Total Population Register. We chose to limit the study to Swedish born children in order to reduce possible ethno-cultural influences among young immigrants’ childbearing and to avoid problems with missing data for parents of immigrants. Cohort youths who had received registered in-home or out-of-home care before age of majority (18 years) were identified in the Swedish Register of Children and Young Persons Subjected to Child Welfare Measures. This is a national register containing mainly data on all out-of-home care placements since the 1970s. The only form of in-home care intervention that can be traced in this database is having been assigned a ‘contact family’, (Andersson & Arvidsson, 2001; Vinnerljung & Franze´n, 2005). This is an intervention defined in Swedish child welfare legislation, where the child welfare authority commissions a voluntary family to provide support and respite care to a young children living in a vulnerable family, or to an older child with behavioural problems (Andersson, 1993; Andersson & Arvidsson, 2001; Barth, 1991). ‘Out-of-home-care’ is foster family and residential care. Residential care of younger children is almost always a short

ARTICLE IN PRESS 102

B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

and temporary affair in Sweden. Intermediate and long-term care, irrespective of age at start, is dominated by placements in foster families. (Vinnerljung, 1996b). Information contained in the database is limited mainly to type of out-of-home care, length of stay and legal ground for the intervention (voluntary placement or court order). No information on the cause for intervention or characteristics of birth parents or children is noted, neither are characteristics of contact families or out-of-home care environments. Forty-nine thousand five hundred and eighty-two youths who had been in in-home or out-ofhome care at some time before age 18 were divided into mutually exclusive subgroups after type of intervention, age at the start of first registered intervention and after time spent in out-of-home care before their 18th birthday (cp. Triseliotis, 1989). Age was defined by birth year and birth month (register data available for this study did not include actual birth day due to confidentiality issues).

 In-home care before teens: Contact family, start of first intervention before age 13 (n ¼ 8698).  Short-term care before teens: 40–o2 years in out-of-home care (foster family and/or residential care), start of first placement before age 13 (n ¼ 9941).

 Intermediate care before teens: 2–o5years in out-of-home care (mostly foster family care, start of first placement before age 13 (n ¼ 2887).

 In-home care during teens: Contact family, start of first intervention at age 13–17 (n ¼ 5472).  Short-term care during teens: o0–o2 years in out-of-home care (foster family and/or residential care), start of first placement at age 13–17 (n ¼ 9772). care during teens: 2–o5 years in out-of-home care (mostly foster family care), start of first placement at age 13–17 (n ¼ 4587). Long-term care: 5–o12 years in out-of home care (mostly foster family care) before 18th birthday, regardless of age at the start of first placement (n ¼ 4740). Grown up in care: Children who spent 12 years or more in out-of-home care (mostly foster family care) (n ¼ 3485).

 Intermediate  

The remaining members of the birth cohorts, without registered experience of in-home or outof-home care before age 18, were labelled the majority population (n ¼ 1; 128; 625). Socio-economic and demographic variables Parents of cohort members were identified in the Swedish Multigenerational Register. We identified 499.9% of the mothers and 99.0% of the fathers of the cohort members. The 1% group with unidentified fathers includes cases where paternal identity has not been legally established. Through the same register, we also identified parents who themselves were teenagers when becoming a mother or father for the first time (before their 20th birthday). We retrieved data on country of birth of cohort members’ birth parents from the Swedish Total Population Register. The Swedish Population and Housing Census of 1980 was the source for the following information about parents of cohort members: SES of the household, housing situation and geographic residency. SES was defined according to a classification by Statistics Sweden, which is based on occupation but also takes educational level, type of position

ARTICLE IN PRESS B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

103

and type of production into account for the head of the household (five categories plus unknown/ missing data). National Censuses were made in 1970, 1980, 1985 and 1990. Variations between censuses in collected information and definitions often complicate retrieving the same kind of data from more than one census. For these reasons, we opted to use the 1980 census for information on SES and demographic status of birth families from cohort members’ early childhood. Using the birth cohorts 1972–1983 for our study, this means that parental data originate from varying ages of the cohort members: from 0 to 8 years after birth for those born 1972–1980 and for individuals born 1981–1983 1–3 years before birth. Outcome variable The Swedish Multigenerational Register was used to establish whether the cohort members had become parents before their 20th birthday (birth of their first child). Statistical analysis Descriptive data are presented in frequency tables, as are unadjusted birth rates. Since the outcome variable is dichotomous and time of follow-up is identical for the entire study population, logistic regression models were used to estimate odds ratios for being a teenage parent. Generally, logistic regression models are used to examine how a specific variable—in a statistical model with several variables—affects the probability of a discrete outcome, given that the values of the other variables are kept constant (for a description of logistic regression analysis, see, e.g. Tabachnick & Fidell, 1996). In model 1, results were adjusted for cohort members’ sex and birth year (entered into the model as a continuous variable). In model 2, we also adjusted for the effects of:

 Birth mother’s and birth father’s country of birth (Sweden, other Nordic countries (Denmark, Norway, Finland and Iceland), other countries and missing data). mother’s and birth father’ region of residency 1980 (city (Stockholm, Gothenburg or Malmo¨), other urban community, mainly rural community or unknown/missing data). Birth mother’s and birth father’s type of housing 1980 (owns house, owns apartment, rents apartment/house or unknown/missing data). Birth mother’s and birth father’s SES in 1980. Familial influences: Birth mother parent before 20th birthday; ditto for birth father.

 Birth   

All analyses were performed in SAS 8.02.

Results Descriptive traits of the child welfare subgroups are listed in Table 2. In comparison with majority population peers, parents of child welfare clients were more likely to have considerably

104

Table 2 Descriptive characteristics of study population, after subgroup

Birth mothers Birth country Sweden Nordic Other countries Missing data Residency City Other urban Mainly rural Unknown/missing data SES Self-employed Non-skilled blue collar Skilled blue collar Lower white collar Medium white collar Higher white collar Unclassified/missing data

In-home care Short-term during teens care during teens N ¼ 5472 N ¼ 9772

Intermediate care Long-term during teens care

Grown up in care

Majority population

N ¼ 4587

N ¼ 3485

N ¼ 1; 128; 625

N ¼ 4740

4767 (54.8%) 5257 (52.9%) 1626 (56.3%) 3512 (64.2%) 4683 (47.9%) 3931 (45.2%) 4684 (47.1%) 1261 (43.7%) 1960 (35.8%) 5089 (52.1%) NA 4 6 NA 16

2402 (52.4%) 2553 (53.9%) 1790 (51.4%) 2185 (47.6%) 2187 (46.1%) 1695 (48.6%) 15 8 2

578,011 (51.2%) 550,614 (48.8%) NA

NA

36

NA

3

40

NA

8

93

190

7107 (81.7%) 7634 (76.8%) 2342 (81.2%) 4705 (86.0%) 8006 (81.9%) 968 (11.1%) 1326 (13.3%) 383 (13.3%) 512 (9.4%) 1012 (10.4%) 623 (7.2%) 954 (9.6%) 160 (5.5%) 255 (4.7%) 753 (7.7%) — 27 (0.1%) 2 (0.1%) — 1 (0.01%)

3795 (82.7%) 3986 (84.1%) 2942 (84.4%) 1,002,305 (88.8%) 503 (11.0%) 580 (12.2%) 435 (12.5%) 71,631 (6.4%) 289 (6.3%) 173 (3.7%) 106 (3.0%) 53,870 (4.8%) — 1 (0.02%) 2 (0.1%) 819 (0.1%)

2795 4436 1314 153

1720 2170 623 74

(32.1%) 3531 (35.5%) 1148 (39.8%) 1350 (24.7%) 3571 (36.5%) (51.0%) 4812 (48.4%) 1302 (45.1%) 2958 (54.1%) 4657 (47.7%) (15.1%) 1223 (12.3%) 361 (12.5%) 1109 (20.3%) 1378 (14.1%) (1.8%) 375 (3.8%) 76 (2.6%) 55 (1.0%) 166 (1.7%)

74 (0.9%) 108 (1.1%) 2604 (30.0%) 2799 (28.2%) 406 (4.7%) 675 (7.8%) 362 (4.2%)

429 (4.3%) 776 (7.8%) 337 (3.4%)

21 (0.7%) 86 (1.6%) 154 (1.6%) 768 (26.6%) 1632 (29.8%) 2839 (29.1%) 106 (3.7%) 153 (5.3%) 65 (2.3%)

272 (4.9%) 477 (8.7%) 274 (5.0%)

506 (5.2%) 915 (9.4%) 510 (5.2%)

105 (1.2%) 95 (1.0%) 18 (0.6%) 80 (1.5%) 136 (1.4%) 4472 (51.3%) 5397 (54.3%) 1756 (60.8%) 2651 (48.5%) 4712 (48.2%)

(37.5%) 1768 (37.3%) 1384 (39.7%) (47.3%) 2218 (46.8%) 1570 (45.1%) (13.6%) 616 (13.0%) 376 (10.8%) (1.6%) 138 (2.9%) 155 (4.5%)

53 (1.2%) 46 (1.0%) 1390 (30.3%) 1145 (24.2%) 207 (4.5%) 414 (9.0%) 180 (3.9%)

140 (3.0%) 240 (5.1%) 83 (1.8%)

319,905 566,654 225,320 16,746

(28.3%) (5 .2%) (20.0%) (1.5%)

18 (0.5%) 686 (19.7%)

32,048 (2.8%) 240,815 (21.3%)

78 (2.2%) 106 (3.0%) 18 (0.5%)

62,201 (5.5%) 166,675 (14.8%) 152,357 (13.5%)

38 (0.8%) 23 (0.5%) 4 (0.1%) 2305 (50.2%) 3 063 (64.6%) 2575 (73.9%)

47,205 (4.2%) 427,324 (37.9%)

ARTICLE IN PRESS

Cohort members Male Female Median age at first placement Median months in out-of-home care before age 18

Intermediate care before teens N ¼ 2887

B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

In-home care Short-term before teens care before teens N ¼ 8698 N ¼ 9941

Housing Owns house Owns apartment Rental housing Unknown/missing data

943 361 3015 268

(20.6%) 570 (12.0%) 205 (5.9%) (7.9%) 299 (6.3%) 197 (5.6%) (65.7%) 3441 (72.6%) 2568 (73.7%) (5.8%) 430 (9.1%) 515 (14.8%)

608,720 92,073 376,326 51,506

(53.9%) (8.2%) (33.3%) (4.6%)

6587 887 1020 204

(75.7%) 6894 (69.4%) 2082 (72.1%) 4510 (82.4%) 7603 (77.8%) (10.2%) 1097 (11.0%) 332 (11.5%) 462 (8.4%) 896 (9.2%) (9.8%) 1299 (13.1%) 284 (9.8%) 423 (9.9%) 1090 (11.6%) (2.4%) 651 (6.6%) 189 (6.6%) 77 (1.4%) 183 (1.9%)

3566 436 453 132

(77.7%) 3679 (77.6%) 2598 (74.6%) (9.5%) 490 (10.3%) 379 (10.8%) (9.9%) 323 (13.1%) 207 (11.2%) (2.9%) 248 (5.2%) 301 (8.6%)

991,361 58,927 70,915 7422

(87.8%) (5.2 %) (6.3%) (0.7%)

2682 4174 1357 485

(30.8%) (48.0%) (15.6%) (5.6%)

1622 2076 628 261

(35.4%) 1588 (33.5%) 1188 (34.1%) (45.3%) 2105 (44.4%) 1442 (41.4%) (13.7%) 637 (13.4%) 375 (10.8%) (5.7%) 410 (8.7%) 480 (13.8%)

316,488 560,531 225,513 26,093

(28.0%) (49.7%) (20.0%) (2.3%)

3296 4343 1241 1061

(33.2%) 1045 (36.2%) 1305 (23.9%) 3444 (35.2%) (43.7%) 1201 (41.6%) 2864 (52.3%) 4476 (45.8%) (12.5%) 356 (12.3%) 1095 (20.0%) 1395 (14.3%) (10.7%) 285 (9.9%) 208 (3.8%) 457 (4.7%)

SES Self-employed Non-skilled blue collar Skilled blue collar Lower white collar Medium white collar Higher white collar Unclassified/missing data

438 (5.0%) 588 (%) 2730 (31.4%) 2728 (%)

120 (4.2%) 419 (7.7%) 653 (%) 807 (27.6%) 1727 (31.6%) 2997 (%)

1715 (19.7%) 1704 (%) 415 (4.8%) 457 (%) 471 (5.4%) 607 (%)

475 (16.5%) 1282 (23.4%) 2031 (%) 100 (3.5%) 330 (6.0%) 603 (%) 110 (3.8%) 372 (6.8%) 719 (%)

Housing Owns house Owns apartment Rental housing Unknown/missing data

1729 691 5159 1119

213 (2.5%) 255 (%) 40 (1.4%) 173 (3.2%) 363 (%) 2716 (31.2%) 3602 (36.2%) 1235 (42.8%) 1169 (21.4%) 2406 (24.6%)

254 (5.5%) 29 (4.4%) 1417 (30.9%) 1440 (30.4%)

108 (3.1%) 906 (26.0%)

108,293 (%) 242,141 (21.5%)

410 (11.8%) 66 (1.9%) 40 (1.2%)

247,852 (22.0%) 95,944 (8.5%) 195,924 (17.4%)

134 (2.9%) 77 (1.6%) 14 (0.4%) 1314 (28.6%) 1973 (41.6%) 1941 (55.7%)

129,456 (11.5%) 109,015 (9.6%)

969 (21.1%) 231 (5.0%) 268 (5.8%)

721 (15.2%) 168 (3.5%) 152 (3.2%)

(19.9%) 1973 (19.8%) 405 (14.0%) 1718 (31.4%) 2520 (25.8%) 1065 (23.2%) 720 (15.2%) 315 (9.0%) (7.9%) 699 (7.0%) 182 (6.3%) 427 (7.8%) 788 (8.0%) 376 (8.2%) 280 (5.9%) 204 (5.8%) (59.3%) 5382 (54.1%) 1764 (61.1%) 2793 (51.0%) 5368 (54.9%) 259,056.5(%) 2822 (59.5%) 2046 (58.7%) (12.9%) 1887 (19.0%) 536 (18.6%) 534 (9.8%) 1106 (11.3%) 556 (12.1%) 918 (19.4%) 920 (26.4%)

Familial influences Birth mother 3462 (34.8%) 1066 (36.9%) 1702 (35.9%) 2820 (28.9%) 1508 (32.9%) teenage parent Birth father teenage 756 (7.6% 256 (8.9%) 400 (8.4%) 659 (6.7%) 324 (7.1%) parent

1328 (38.1%) 2358 (29.2%) 1591 (29.1%) 342 (9.8%)

568 (6.5%)

363 (6.6%)

619,311 88,508 355,026 65 780

(54.9%) (7.8%) (31.5%) (5.8%)

149,321 (13.2%) 30,692 (2.7%)

ARTICLE IN PRESS

Residency City Other urban Mainly rural Unknown/missing data

(16.8%) 1627 (16.4%) 362 (12.5%) 1603 (29.3%) 2340 (24.0%) (8.6%) 788 (7.9%) 174 (6.0%) 462 (8.4%) 843 (8.6%) (68.4%) 6647 (66.9%) 2082 (72.1%) 3120 (57.0%) 5965 (61.0%) (6.2%) 879 (8.8%) 269 (9.3%) 287 (5.25%) 624 (6.4%)

B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

Birth fathers Birth country Sweden Nordic Other countries Missing data

1465 751 5947 535

105

ARTICLE IN PRESS 106

B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

lower SES, to rent instead of own their house or apartment and to live in cities. Also, more were teenage parents themselves, both mothers and fathers. Of all teenage parents in the cohorts, almost every fifth (18%) is a former child welfare client. Crude prevalence rates of teenage births after subgroups are listed in Table 3. Among youth who received interventions first during their teens, 16–19% of the girls gave birth to a child before age 20, compared to 2.9% in the majority population, for boys figures varied between 5.0 and 5.6% compared to 0.7% among majority population peers. For the other child welfare groups, including children in long-term foster family care, rates for girls were 9–12% and for boys 2.4–4.2%. Looking at teenage parent prevalence figures for cohort members and their birth parents, rates reflect a radical change in population teenage birth patterns over time. In the majority population group, 13.2% of the mothers to cohort members were teenage parents and 2.7% of the fathers (end of Table 2). Rates in the birth cohorts born 1972–1983 were 2.9 and 0.7% (Table 3), in comparison to about 15 for girls and 14 for boys. Results from logistic regression analysis are summarized in Table 4. When results were adjusted for sex and birth year (model 1), odds for youth who received interventions during their teens were six- to ninefold in comparison with general population peers. For other groups, odds varied between three- and fivefold. After adjusting for socio-economic and demographic background and for familial influences, odds ratios were lowered to about four- to fivefold for the teenage groups and two- to threefold for the others, except for children who grew up in care. The latter group, where the median individual spent almost 16 years in out-of-home care before age of majority, had only a moderately elevated odds ratio, 1.5 (95% confidence interval (CI) 1.3–1.7). Analyses performed on girls and boys separately yielded only minor differences between sexes, but produced odds ratios with wide CIs for boys, due to few cases in each subgroup. The well-known trend of familial influence on teenage parenthood, prominent in the majority population, was present for all child welfare groups after adjusting for socio-economic background of birth home, irrespective of length of stay in substitute care (Table 5). Familial factors tended to be somewhat weaker for the subgroups that received services in adolescent years, possibly an effect of small sample size in subgroups, evident by wide CIs. Again, separate analyses for boys and girls did not yield substantial variations between sexes, when broken down into subgroups. Finally, a note on the alleged ‘‘cycle of teenage parenthood’’, a popular theory among child welfare practitioners in some countries (see, e.g. Rutman, Strega, Callahan, & Dominelli, 2002). Although odds for becoming a teenage parent were elevated for youth with familial influences of teenage parenthood, the large majority of these cohort members did not repeat their parents’ pattern (results from bivariate analysis not shown in tables). For females with a teenage birth mother, 7% in the majority population became teenage mothers themselves. In the child welfare groups, rates were considerably higher: 21–27% for the groups that had received interventions during their teens and 13–23% for the others. The same trend was valid for cohort members with teenage birth fathers. Among majority population cohort girls with a teenage father, 8% became teenage parents, 17–23% among girls in the teenage intervention groups and 16–18% in the other child welfare subgroups. Results for boys had a similar gradient for familial influences from both the birth mother and the birth father.

Table 3 Prevalence of teenage parents, after subgroup Short-term care Intermediate In-home care before teens care before teens during teens

N ¼ 8698

N ¼ 9941

N ¼ 2887

N ¼ 3485

N ¼ 1; 128; 625

360 (9.2%) 112 (2.4%)

538 (11.5%) 126 (2.4%)

196 (15.4%) 69 (4.2%)

307 (15.7%) 175 (5.0%)

976 (19.2%) 245 (5.2%)

367 (16.8%) 134 (5.6%)

158 (9.3%) 49 (2.7%)

15,293 (2.9%) 4164 (0.7%)

Intermediate Long-term Grown up in Majority care during care care population teens OR (95% CI) OR (95% CI) OR (95% CI) OR (reference) 7.6 (6.9–8.4) 4.1 (3.7–4.5)

259 (11.8%) 83 (3.3%)

Table 4 Summary of logistic regressions for becoming teenage parent, after subgroup

Model 1 Model 2

In-home care before teens

Short-term care Intermediate In-home care before teens care before teens during teens

OR (95% CI)

OR (95% CI)

OR (95% CI)

OR (95% CI)

Short-term care during teens OR (95% CI)

4.0 (3.6–4.4) 2.1 (1.9–2.3)

4.1 (3.7–4.4) 2.1 (1.9–2.3)

6.5 (5.7–7.4) 2.9 (2.6–3.4)

6.9 (6.3–7.6) 4.2 (3.8–4.6)

8.8 (8.2–9.4) 5.4 (5.1–5.8)

4.7 (4.2–5–2) 3.6 (3.1–4.2) 2.1 (1.9–2.4) 1.5 (1.3–1.7)

1 1

Model 1 is adjusted for cohort member’s year of birth and sex. Model 2 is adjusted for cohort member’s year of birth and sex and mother’s/father’s birth country, residency, form of housing, SES and familial influences (mother or father a teenage parent).

ARTICLE IN PRESS

Intermediate Long-term care during care teens N ¼ 4587 N ¼ 4740

Grown up in Majority care population

N ¼ 5472

Short-term care during teens N ¼ 9772

B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

Cohort girls Cohort boys

In-home care before teens

107

108

Short-term care Intermediate In-home care before teens care before teens during teens OR (95% CI)

Short-term care during teens OR (95% CI)

Intermediate Long-term Grown up in Majority care during care care population teens OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)

OR (95% CI)

OR (95% CI)

OR (95% CI)

2.2 (1.8–2.6)

1.7 (1.5–2.1)

2.2 (1.6–2.8)

1.4 (1.1–1.7)

1.6 (1.4–1.8)

1.6 (1.3–1.9) 2.0 (1.6–2.5)

1.6 (1.1–2.2)

2.3 (2.2–2.3)

1.7 (1.2–2.3)

[1.1 (0.8–1.4)]

[1.1 (0.7–1.9)]

[1.4 (1.0–2.0)]

[1.2 (0.9–1.5)] [0.9 (0.6–1.2)] 1.5 (1.1–2.1)

1.7 (1.2–2.6)

1.6 (1.5–1.7)

Model 2 is adjusted for cohort member’s year of birth and sex and mother’s/father’s birth country, residency, form of housing, SES and familial influences (mother or father a teenage parent). [OR] ¼ statistically non-significant (pX0.05).

ARTICLE IN PRESS

Birth mother teenage parent Birth father teenage parent

In-home care before teens

B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

Table 5 Summary of logistic regressions. Interaction effects of familial influences with subgroup

ARTICLE IN PRESS B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

109

Discussion This is, to our knowledge, the first large-scale population-based study examining the prevalence and odds for teenage parenthood among child welfare clients that discriminates between subgroups within the child welfare population and adjusts results for socio-demographic and familial background factors. The results confirm, moderate and expand results from other studies, showing high prevalence figures and highly elevated odds in some subgroups. But we found substantial variation within the child welfare population, trends that remained after adjusting results for a number of background factors. As could be expected from a review of the literature, youth who became child welfare clients in their teens had strikingly high prevalence figures when compared to other cohort peers. Among girls, one in five or one in six had given birth in their teens in contrast to about one in 35 in the majority population. About one in 20 among boys had become a teenage father, compared to one in 140 among peers without registered experience of child welfare interventions. Odds for becoming a teenage parent in these subgroups were six- to ninefold, compared to majority population peers, when adjusting for sex and birth year only. These odds ratios were lowered when we adjusted for socio-economic, demographic and familial factors in the logistic regression models, but remained four- to sixfold. For cohort members whose interventions started before age 13, and for youth who had been in long-term foster care, roughly 10% of the girls and 2–4% of the boys had become teenage parents. After adjusting for birth family-related background factors, odds ratios were mostly twofold. Youth who grew up in foster homes (median 16 years in care) had only moderately elevated odds (OR ¼ 1.5). The influence of familial factors on teenage parenthood, well known in the literature (e.g. Olausson, Lichtenstein, & Cnattingius, 2000), was also present for the child welfare client groups, after we adjusted results for socio-economic background factors. But when looking at crude prevalence rates, the great majority of cohort members with a teenage mother or father did not become teenage parents themselves. As Michael Rutter has stressed in many of his works on human development over the life course; whenever we see continuity between generations, we also find more discontinuity (e.g. Rutter, 1998). A national study on adolescent out-of-home placements found that 80% of the boys and 60% of the girls were described in case files as having behavioural problems, mostly antisocial behaviour and/or advanced absconding from school (Vinnerljung, Sallna¨s, & Kyhle Westermark, 2001). Results from our study on the adolescent intervention groups should be viewed in this perspective, probably verifying well-known associations between antisocial behaviour and teenage parenthood for both girls and boys (e.g. Rutter et al., 1998). But available register data do not enable us to identify characteristics of individual children, thereby excluding the possibility of a more precise analysis. Antisocial youth of both sexes tend to have a behavioural pattern of risk-taking behaviour. Liaisons and assorted mating with antisocial and risk-taking partners are common, as is a casual attitude to the use and maintenance of birth control (e.g. Corlyon & McGuire, 1999; Eke´us & Christensson, 2003; Holmberg & Berg-Kelly, 2002; Moffit, Caspi, Rutter, & Silva, 2001; Rutter, Giller & Hagell, 1998). Interviews with 515 12- to 20-year-old youth in Swedish secure units using the ADAD scale (Adolescent Drug Abuse Diagosis; Friedman & Utada, 1989) showed that every

ARTICLE IN PRESS 110

B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

third girl had been pregnant at least once and almost as many boys admitted to having made a girl pregnant (cp. Dennison & Lyon, 2001; SiS, 2005; cp. Dennison & Lyon, 2001). In many Western countries, ‘‘conventional wisdom’’ claims that many single and poor young women choose deliberately to have children in adolescence, for financial reasons or to assign themselves a tangible role in a society where they otherwise are marginalized. Research from several nations does not support that notion. The vast majority of teenage pregnancies are unplanned and perceived by the young women themselves as accidental (e.g. review in Social Exclusion Unit, 1999). But British research also suggests that young women in out-of-home care may be more likely than their average peers to reject abortions on ethical grounds—like many other disadvantaged adolescents (Smith, 1993; Turner, 2004)—and go through with an unplanned pregnancy (Corlyon & McGuire, 1999). Also, one study indicated that both girls and boys in this group have more positive attitudes towards early childbearing than peers. These attitudes, coupled with a common fatalistic view of a future they feel they cannot control, seem to contribute to girls’ acceptance of unintended teenage pregnancies (Corlyon & McGuire, 1999). Whether this is also valid for Sweden, we do not know due to the lack of similar studies. Prevalence figures and adjusted odds ratios were elevated—more or less—for all child welfare groups that received service before teens. Looking at youth who spent almost their entire formative years in foster homes, we found that they still had poorer outcome than peers from families with similar socio-economic and familial background as the youths’ birth families. In comparison with the general population, more Swedish foster parents have only primary education, work blue-collar jobs and live in semi-rural or rural settings (Ho¨jer, 2001). But they hold distinctly better socio-economic positions than the birth parents of children in out-of-home care (Hessle, 1988; Vinterhed, 1985). It is not possible to trace foster families in Swedish national registers, and henceforth not possible to adjust results for background variables related to the families where the long-term foster care children actually grew up. Our study has no intention of evaluating child welfare interventions, and we cannot estimate how the absence of, for example, placements in out-of-home care would affect prevalence rates and odds. But the results do not suggest any radical effects of these interventions, not even when they lasted many years. As for child welfare interventions targeting adolescents, we know that these youths tend to be carriers of many ‘‘risk factors’’, often appearing in clusters (Rutter & Maugham, 1997; Stattin & Magnusson, 1996). Looking at our results, a pivotal question seems inescapable: do child welfare interventions produce protective processes that reduce the risk of teenage parenthood? Do some interventions even have harmful effects that increase the risk of teenage pregnancy and parenthood, for example, placement of girls with behaviour problems in group care (cp. Andreassen, 2003; Dishion, McCord, & Poulin, 1999)? Preliminary results from an on-going follow-up study at age 25 of over 700 youths placed in out-of-home care in 1991 do not seem to support the hypothesis that girls in Swedish residential care have higher risks for becoming teenage parents than girls placed in foster care. Instead, the dividing line seems to be placement for antisocial behavioural or not. Just over 25% of 130 girls, placed for reasons related to antisocial behaviour, became teenage parents, the same proportion among those placed in foster families as in residential care. Rates were considerably lower for girls entering care for other reasons, regardless of care setting (Vinnerljung & Sallna¨s, forthcoming). Corlyon and McGuire

ARTICLE IN PRESS B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

111

(1999) found that other factors than the type of out-of-home care seemed to have detrimental influences. Girls in care were, for example, less likely than peers to have completed sex and birth control education in school, and usually their carers—foster parents and residential care workers—did not provide the girls with compensatory help. In the absence of guidelines from local authorities, provision of information and support on vital issues like birth control and abortion became individual ad hoc choices for carers, often basing their decisions on moral preferences.

Conclusions and implications Causality issues aside, youth who become child welfare clients in adolescence should be regarded as a high-risk group for teenage parenthood, regardless if they receive in-home or out-ofhome care. Concern is also valid for those who enter child welfare services in younger years and/ or spend a large part of their childhood in long-term foster care, albeit to a lesser degree. Even though elevated rates and odds for some child welfare client groups can in part be statistically explained by socio-demographic background factors related to their birth families, crude prevalence figures several times higher than majority population peers remain, and indicate needs of preventive services. Considering that youth who enter the child welfare system in adolescence have the highest risks during the years they are objects of interventions, and often are in the physical custody of the child welfare authorities, there ought to be ample opportunities for preventive action. Meta-analyses, based mainly on studies from the US, indicate that the most efficient teenage pregnancy prevention programmes for girls combine elements of education and distribution of contraceptives (Ferrer-Wreder, Stattin, Lorente, Tubman, & Adamson, 2004; Franklin, Grant, Corcoran, O’Dell Miller, & Bultman, 1997; see also Peckham, 1993). Subsequently, agencies responsible for adolescent girls in care settings should, as a minimum, provide every individual youth with access to clinics that offer such services. Also, it may be important to monitor birth control maintenance, given reports on casual attitudes on such matters among girls in out-of-home care (Corlyon & McGuire, 1999). Adolescent boys involved in pregnancies have been objects of far less attention than girls, both in clinical practice and scientific studies (Holmberg & Berg-Kelly, 2002). Since boys and girls who enter the child welfare system tend to share the same backgrounds and develop the same excess risks of becoming teenage parents, they ought to be targeted by similar ambitions of service provision. But improving access to mainstream services may not be enough. Research aiming to increase our limited understanding of reproductive behaviour among troubled adolescents holds a great potential for clinical relevance. At present, we do not even know if the high teenage birth rates found in our study is due to more to higher pregnancy rates or to lower abortion rates, or an even combination of both. Can mainstream services be adapted to the needs and preferences of antisocial youth, or are they better served by (theory-based) intervention programmes, specifically targeting high-risk groups? If the latter, can existing evidence-based prevention programmes, proven to be effective for high risk youth (e.g. the US Teen Outreach Program; Allen & Philliber, 2001; Allen, Philliber, Herrling, & Kuperminc, 1997), be adapted to, for example, out-of-home care populations? As far as we know, this has not been tried.

ARTICLE IN PRESS 112

B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

Limitations Large-scale register studies have inherent limitations, and ours is no exception. Firstly, all child welfare interventions cannot be traced in the national registers. Also, there are known, minor problems with under-reporting from local authorities of those interventions that should be registered (e.g. Vinnerljung & Ribe, 2001). Subsequently, nested within the majority population is a group of youths who has received other forms of in-home care than a ‘‘contact family’’ or simply have not been reported by the local authorities to the national database. Even if these errors are large enough to affect results—which is not likely due to the large sample size of majority population peers—they would cause only minor underestimations of odds ratios for the child welfare subgroups. A brief note on adoptions: a subgroup of international adoptees (about 1–1.5% of these birth cohorts, mostly adopted in infancy) is also nested within the study groups, almost all within the majority population (Hjern, Lindblad, & Vinnerljung, 2002; Vinnerljung, 1999). Their adoptive parents are treated as birth parents in the analysis. In contrast to the US and the UK, Swedish legislation does not permit national adoption without birth parents’ consent, and only a handful of national adoptions are passed through the courts yearly (Barth, 1992; Lindblad, 2004). Secondly, the registers contain mostly data from administrative processes in different ‘‘social control agencies’’. We have no data on childhood aggression, antisocial behaviour, parental warmth in birth homes, etc. that would permit construction of theory-based models. This study can therefore not go beyond explorative and descriptive aims. Also, due to the limitations of the national register containing child welfare data, we are unable to differentiate youth who received interventions for antisocial behaviour problems from others, thereby limiting our analysis. Thirdly, data on socio-economic backgrounds of the cohort members were mainly retrieved from the National Census, 1980, as discussed in the Method section. This selection is not optimal, but a ‘‘choice of lesser evils’’ due to problems with varying definitions in census data over time. We have also omitted family status (single, married, single adult household, etc.) from our analysis for two reasons. Data on legal civil status (married, non-married, divorced) tend to be less meaningful (especially for groups in the social marginal) since many Swedish households consist of unmarried couples with children. Census data are also generally considered of questionable reliability for the purpose of assessing single adult household status, often a transient condition for women in child-bearing age (especially in lower SES groups). Finally (as has been mentioned before), we lack data on substitute care families, making it impossible to adjust results for socio-economic conditions of the environment where children in long-term foster care actually grew up. Acknowledgements Sincere thanks to our colleagues Petra Otterblad Olausson and Anders Hjern, at the Center for Epidemiology, for generously sharing their knowledge and wisdom. References Allen, J., & Philliber, S. (2001). Who benefits most from a broadly targeted prevention program? Differential efficacy across populations in the Teen Outreach Program. Journal of Community Psychology, 29, 637–655.

ARTICLE IN PRESS B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

113

Allen, J., Philliber, S., Herrling, S., & Kuperminc, G. P. (1997). Preventing teen pregnancy and academic failure: experimental evaluation of a developmentally based approach. Child Development, 64, 729–742. Andersson, G. (1993). Support and relief: The Swedish contact person and contact family program. Scandinavian Journal of Social Welfare, 2, 54–62. Andersson, G., Arvidsson, M. B., (2001). Vad vet vi om insatsen kontaktperson-familj. En kunskapso¨versikt [What do we know about contact persons/-families? A review.] Lunds: Universitet, Meddelanden fra˚n Socialho¨gskolan, publ. no. 1. Andreassen, T. (2003). Institutionsbehandling av ungdomar. Vad sa¨ger forskningen? [Residential treatment of adolescents. A review.]. Stockholm: Gothia/CUS/SiS. Barth, R. (1990). On their own: The experiences of youth after foster care. Child and Adolescent Social Work Journal, 7, 419–440. Barth, R. (1991). Sweden’s Contact Family Program. Informal help bolsters vulnerable families. Public Welfare, 49, 36–46. Barth, R. (1992). Child Welfare services in the United States and Sweden: Different assumptions, laws and outcomes. Scandinavian Journal of Social Welfare, 1, 36–42. Bebbington, A., & Miles, J. (1989). The background of children who enter local authority care. British Journal of Social Work, 19, 349–368. Biehal, N., Clayden, J., Stein, M., & Wade, J. (1992). Prepared for living?. London: NCB. Biehal, N., Clayden, J., Stein, M., & Wade, J. (1995). Moving on. Young people and leaving care schemes. London: HMSO. Botting, B., Rosato, M., & Wood, R. (1998). Teenage mothers and the health of their children. Population Trends, 93, 19–28. Buchanan, A. (1999). Are care leavers significantly dissatisfied and depressed in adult life? Adoption & Fostering, 23, 35–40. Bunting, L., & McAuley, C. (2004). Research review. Teenage pregnancy and parenthood: the role of fathers. Child and Family Social Work, 9, 295–303. Cashmore, J., & Paxman, M. (1996). Wards leaving care: A longitudinal study. NSW, Sydney: Department of Community Services. Chilman, C. S. (1980). Social and psychological research concerning adolescent childbearing 1970–1980. Journal of Marriage and the Family, 42, 793–805. Christoffersen, M. N. (1993). Anbragte bo¨rns livsfo¨rlo¨p [The life course of children from out-of-home care]. Copenhagen: Socialforskningsinstituttet rapport 93, publ. no. 11. Christoffersen, M. N. (2005). Motherhood and induced abortion among teenagers: A longitudinal study of all 15–19 year old Danish women born in 1966. Paper presented at the 15th Nordic demographic symposium, April 28–30, 2005, A˚lborg, Denmark. Coley, R. L., & Chase-Lansdale, P. L. (1998). Adolescent pregnancy and parenthood. Recent evidence and future directions. American Psychologist, 53, 152–166. Cook, R. (1994). Are we helping foster care youth prepare for their future? Children and Youth Services Review, 16, 213–229. Corlyon, J., & McGuire, C. (1999). Pregnancy and parenthood—The views and experiences of young people in public care. London: NCB. Courtney, M., & Piliavin, I. (1998). Foster youth transitions to adulthood: Outcomes 12–18 months after leaving out-ofhome care. Madison, WI: University of Wisconsin. Danielsson, M., Rogala, C., & Sundstro¨m, K. (2001). Teenage sexual and reproductive behavior in developed countries: Country report for Sweden. New York: The Guttmacher Institute Occasional Report # 7. Danielsson, M., Rogala, C., & Sundstro¨m, K. (2003). Fa˚ tona˚rsgraviditeter i Sverige—ja¨mfo¨relse mellan fem va¨stla¨nder [Few teenage pregnancies in Sweden—a comparison between five Western countries]. La¨kartidningen, 100, 2063–2066. Darroch, J. E., Frost, J. J., & Singh, S. (2001). Teenage sexual and reproductive behavior in developed countries: Can more progress be made?. New York: The Alan Guttmacher Institute Occasional Report #3. Dennison, C., & Lyon, J. (2001). Young offenders, fatherhood and the impact of parenting training. Brighton: Trust for the Study of Adolescence.

ARTICLE IN PRESS 114

B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

Dishion, T., McCord, J., & Poulin, F. (1999). When interventions harm. Peer groups and problem behaviors. American Psychologist, 54, 755–764. Dixon, L., Browne, K., & Hamilton-Giachritsis, C. (2005). Risk factors of parents abused as children: A mediational analysis of the intergenerational continuity of child maltreatment (Part I). Journal of Child Psychology and Psychiatry, 46, 47–57. Eke´us, C., & Christensson, K. (2003). Socioeconomic characteristics of fathers of children born to teenage mothers in Stockholm, Sweden. Scandinavian Journal of Public Health, 31, 73–76. Fagot, B., Pears, K., Capaldi, D., Crosby, L., & Leve, C. (1998). Becoming an adolescent father: Precursors and parenting. Developmental Psychology, 34, 1209–1219. Ferrer-Wreder, L., Stattin, H., Lorente, C. C., Tubman, J., & Adamson, L. (2004). Successful prevention and youth development programs. Across borders. New York: Kluwer Academic Publishers. Franklin, C., Grant, D., Corcoran, J., O’Dell Miller, P., & Bultman, L. (1997). Effectiveness of prevention programs for adolescent pregnancy: A meta-analysis. Journal of Marriage and the Family, 59, 551–567. Fraser, A., Brockert, J., & Ward, R. H. (1995). Association of young maternal age with adverse reproductive outcomes. The New England Journal of Medicine, 332, 1113–1117. Friedman, A., & Utada, A. (1989). A method for diagnosing and planning the treatment for adolescent drug abusers. Journal of Drug Education, 19, 285–312. Furstenberg, F., Brooks-Gunn, J., & Chase-Lansdale, L. (1989). Teenage pregnancy and childbearing. American Psychologist, 44, 313–320. Furstenberg, F., Levine, J., & Brooks-Gunn, J. (1990). The children of teenage mothers: Patterns of early childbearing in two generations. Family Planning Perspectives, 22, 54–61. Garnett, L. (1992). Leaving care and after. London: National Children’s Bureau. Hardy, J., Astone, N., Brooks-Gunn, J., Shapiro, S., & Miller, T. (1998). Like mother, like child: Intergenerational patterns of age at first birth and associations with childhood and adolescent characteristics and adult outcomes in the second generation. Developmental Psychology, 34, 1220–1232. Heesle, S., & Vinnerljung, B. (1999). Child welfare in Sweden—An overview. Stockholm: Stockholm University Stockholm Studies in Social Work, #15. Hjern, A., Lindblad, F., & Vinnerljung, B. (2002). Suicide, psychiatric illness and social maladjustment in intercountry adoptees in Sweden: A cohort study. Lancet, 360, 443–448. Hessle, S. (1988). Familjer i so¨nderfall [Disintegrating families]. Gothenburg: Norstedts. Hobcraft, J., & Kiernan, K. (2001). Childhood poverty, early motherhood and adult social exclusion. British Journal of Sociology, 52, 495–517. Ho¨jer, I. (2001). Fosterfamiljens inre liv [Inner life of foster families]. Dissertation, Gothenburg: Department of Social Work, University of Gothenburg. Holmberg, L., & Berg-Kelly, K. (2002). Health, health-compromising behaviour, sexuality and involvement in pregnancy among 18-year-old males: A cross sectional survey. Acta Paediatrica, 91, 838–843. Irvine, H., Bradley, T., Cupples, M., & Boohan, M. (1997). The implications of teenage pregnancy and motherhood for primary health care: Unresolved issues. British Journal of General Practice, 47, 323–326. Jones, E., Forrest, J. D., Goldman, N., Henshaw, S., Lincoln, R., Rosoff, J., Westoff, C., & Wulf, D. (1985). Teenage pregnancy in developed countries: Determinants and policy implications. Family Planning Perspectives, 17, 53–63. Kaufman, J. (1996). Teenage parents and their offspring. Annals of the New York Academy of Sciences, 789, 17–30. Kiernan, K. E. (1997). Becoming a young parent: A longitudinal study of associated factors. British Journal of Sociology, 48, 406–427. Lee, B. J., & Goerge, R. M. (1999). Poverty, early childbearing, and child maltreatment: A multinomial analysis. Children and Youth Services Review, 21, 755–780. Lewinsohn, P., Rohde, P., & Seeley, J. (1994). Psychosocial risk factors for future adolescent suicide attempts. Journal of Consulting and Clinical Psychology, 62, 297–305. Lindblad, F. (2004). Adoption. Lund: Studentlitteratur. Manlove, J. (1997). Early motherhood in an intergenerational perspective: The experiences of a British cohort. Journal of Marriage and the Family, 59, 263–279.

ARTICLE IN PRESS B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

115

Maskey, S. (1991). Teenage pregnancy: Doubts, uncertainties and psychiatric disturbance. Journal of the Royal Society of Medicine, 84, 723–725. McMillen, J. C., & Tucker, J. (1999). The status of older adolescents at exit of out-of-home care. Child Welfare, 78, 339–360. Mendes, P., & Moslehuddin, B. (2004). Graduating from the child welfare system: A comparison of the UK and Australian leaving care debates. International Journal of Social Welfare, 13, 332–339. Miller, B., & Moore, K. (1990). Adolescent behavior, pregnancy, and parenting: Research through the 1980s. Journal of the Marriage and the Family, 52, 1025–1044. Mittendorfer-Rutz, E., Rasmussen, F., & Wasserman, D. (2004). Restricted fetal growth and adverse maternal psychosocial and socioeconomic conditions as risk factors for suicidal offspring: A cohort study. Lancet, 364, 1135–1139. Moffit, T., Caspi, A., Rutter, M., & Silva, P. (2001). Sex differences in antisocial behaviour. Cambridge: Cambridge University Press. OECD. (2002). Society at a glance—www/oecd.org/els/social/indicators Olausson, P. O., Cnattingius, S., & Goldenberg, R. (1997). Determinants of poor pregnancy outcomes among teenagers in Sweden. Obstetrics and Gynecology, 89, 451–457. Olausson, P. O., Cnattingius, S., & Haglund, B. (1999). Teenage pregnancies and risk of late fetal death and infant mortality. British Journal of Obstetrics and Gynaecology, 106, 116–121. Olausson, P. O., Haglund, B., Weitoft, G. R., & Cnattingius, S. (2001). Teenage childbearing and long-term socioeconomic consequences: A case study in Sweden. Family Planning Perspectives, 33, 70–74. Olausson, P. O., Haglund, B., Weitoft, G. R., & Cnattingius, S. (2004). Premature death among teenage mothers. BJOG: An International Journal of Obstetrics and Gynaecology, 111, 793–799. Olausson, P. O., Lichtenstein, P., & Cnattingius, S. (2000). Aetiology of teenage childbearing: Reasons for familial effects. Journal of Twin Research, 3, 23–27. Peckham, S. (1993). Preventing unintended teenage pregnancies. Public Health, 107, 125–133. Polit, D., Morton, T., & White, C. M. (1989). Sex, contraception and pregnancy among adolescents in foster care. Family Planning Perspectives, 21, 203–208. Rutman, D., Strega, S., Callahan, M., & Dominelli, L. (2002). ‘Undeserving’ mothers? Practitioners’ experiences working with young mothers in/from care. Child and Family Social Work, 7, 149–159. Rutter, M. (1998). Some research considerations on intergenerational continuities and discontinuities: Comment on the special section. Developmental Psychology, 34, 1269–1273. Rutter, M., Giller, H., & Hagell, A. (1998). Antisocial behavior by young people. Cambridge: Cambridge University Press. Rutter, M., & Maugham, B. (1997). Psychosocial adversities in childhood and adult psychopathology. Journal of Personality Disorders, 11, 4–18. Santow, G., & Bracher, M. (1999). Explaining trends in teenage childbearing in Sweden. Studies in Family Planning, 30, 169–182. Scaramella, L., Conger, R., Simons, R., & Whitbeck, L. (1998). Predicting risk for pregnancy by late adolescence: A social contextual perspective. Developmental Psychology, 34, 1233–1245. Serbin, L., Cooperman, J., Peters, P., Lehoux, P., Stack, D., & Schwarzman, A. (1998). Intergenerational transfer of psychosocial risk in women with childhood histories of aggression, withdrawal or aggression and withdrawal. Developmental Psychology, 34, 1246–1262. Serbin, L., & Karp, J. (2004). The intergenerational transfer of psychosocial risk: Mediators of vulnerability and resilience. Annual Review of Psychology, 55, 333–363. SiS. (2005). A˚rsrapport ADAD 03 [Annual report ADAD 03]. Stockholm: Allma¨n SiS-rapport 2005, publ. no. 1. Smith, T. (1993). Influence of socioeconomic factors on attaining targets for reducing teenage pregnancies. British Medical Journal, 306, 1232–1235. Social Exclusion Unit. (1999). Teenage pregnancy. Norwich: The Stationary Office. Stattin, H., & Magnusson, M. (1996). Antisocial development: A holistic approach. Development and Psychopathology, 8, 617–645. Sundell, K., Vinnerljung, B., Lo¨fholm, C. A., Humlesjo¨, E. (2004). Social barnava˚rd i Stockholm [Child welfare in Stockholm]. Stockholm: Stockholms stad, FoU-rapport 2004, publ. no. 4.

ARTICLE IN PRESS 116

B. Vinnerljung et al. / Journal of Adolescence 30 (2007) 97–116

Tabachnick, B., & Fidell, L. (1996). Using multivariate statistics (3rd ed). New York: Harper Collins. Triseliotis, J. (1989). Foster care outcomes: a review of key research findings. Adoption & Fostering, 13, 5–17. Turner, K. (2004). Young women’s views on teenage motherhood: A possible explanation for the relationship between socio-economic background and teenage pregnancy outcome? Journal of Youth Studies, 2, 221–238. Vikat, A., Rimpela¨, A., Kosunen, E., & Rimpela¨, M. (2002). Socio-demographic differences in the occurrence of teenage pregnancies in Finland in 1987–1998: A follow up study. Journal of Epidemiology and Community Health, 56, 659–668. Vinnerljung, B. (1996a). Fosterbarn som vuxna [Foster children as adults]. Lund: Arkiv. Vinnerljung, B. (1996b). Svensk forskning om fosterbarnsva˚rd—En o¨versikt [Swedish research on foster care—A review]. Stockholm: Liber Utbildning. Vinnerljung, B. (1999). Fo¨rekomst av adoptivbarn och la˚ngtidsva˚rdade fosterbarn bland placeringar av tona˚ringar i dygnsva˚rd [Prevalence of adoptees and children from long-term foster care in placements of teenagers in care]. Socialvetenskaplig Tidskrift, 6, 313–328. Vinnerljung, B., & Franze´n, E. (2005). Kontaktfamilj/kontaktperson—Omfattning och samband med placering i dygnsva˚rd [Contact family/-person—Prevalence and correlation with out-of-home care]. Socialvetenskaplig Tidskrift, 12, 345–365. Vinnerljung, B., Hjern, A., & Lindblad, F. (in press). Suicide attempts and severe psychiatric morbidity among former child welfare clients—A national cohort study. Journal of Child Psychology and Psychiatry. Vinnerljung, B., O¨man, M., & Gunnarson, T. (2005). Educational attainments of former child welfare clients—A Swedish national cohort study. International Journal of Social Welfare, 14, 265–276. Vinnerljung, B., & Ribe, M. (2001). Mortality after care among young adult foster children in Sweden. International Journal of Social Welfare, 10, 164–173. Vinnerljung, B., & Sallna¨s, M. (forthcoming). Adolescents placed in out-of-home care as young adults—A follow-up at age 25. Vinnerljung, B., Sallna¨s, M., & Kyhle Westermark, P. (2001). Sammanbrott vid tona˚rsplaceringar [Breakdowns in placements of teenagers in out-of-home care]. Stockholm: Socialstyrelsen/CUS. Vinnerljung, B., Sundell, K., Lo¨fholm, C. A., & Humlesjo¨, E. (2006). Former Stockholm child protection cases as young adults: do outcomes differ between those that received services and those that did not? Children and Youth Services Review, 28, 59–77. Vinterhed, K. (1985). De andra fo¨ra¨ldrarna [The other parents]. Stockholm: Skeab. Wellings, K., Wadsworth, J., Johnson, A., Field, J., & MacDowall, W. (1999). Teenage fertility and life chances. Reviews of Reproduction, 4, 184–190. Widom, C. S., & Kuhns, J. (1996). Childhood victimization and subsequent risk for promiscuity, prostitution, and teenage pregnancy: A prospective study. American Journal of Public Health, 86, 1607–1612. Williams, S., Forbes, J. F., McIlwaine, G. M., & Rosenberg, K. (1987). Poverty and teenage pregnancy. British Medical Journal (Clinical research ed.), 294, 20–21.