Early Human Development 100 (2016) 1–5
Contents lists available at ScienceDirect
Early Human Development journal homepage: www.elsevier.com/locate/earlhumdev
The influence of multiple birth and bereavement on maternal and family outcomes 2 and 7 years after very preterm birth Karli Treyvaud a,b,c,d,⁎, Andrea C. Aldana e, Shannon E. Scratch b, Alexandra M. Ure b, Carmen C. Pace b, Lex W. Doyle b,c, Peter J. Anderson b,c a
LaTrobe University, Victoria, Australia Murdoch Childrens Research Institute, Victoria, Australia University of Melbourne, Victoria, Australia d Royal Women's Hospital, Victoria, Australia e Universite Laval, Quebec, Canada b c
a r t i c l e
i n f o
Article history: Received 9 November 2015 Received in revised form 21 April 2016 Accepted 21 April 2016 Keywords: Preterm Maternal depression Maternal anxiety Multiple birth
a b s t r a c t Background: Psychological distress has been reported by mothers of infants born very preterm (VPT) and by mothers of multiples (twins and triplets). This study examined the influence of i) multiple birth and ii) bereavement associated with a multifetal pregnancy, on mental health, parenting stress and family functioning for mothers of children born VPT across early childhood. Methods: Participants were 162 mothers of 194 infants (129 singletons, 65 multiples) born at b 30 weeks' gestation or with a birth weight b 1250 g who completed questionnaires when their children were two and seven years corrected age. Fifteen mothers (9%) experienced bereavement associated with a multifetal pregnancy. Maternal mental health was assessed using the General Health Questionnaire at two years and Hospital Anxiety and Depression Scale at seven years. Parenting stress and family functioning were assessed using the Parenting Stress Index and Family Assessment Device. Results: Maternal mental health, stress and family functioning were similar in mothers of VPT singletons and multiples. However compared with mothers who had not experienced bereavement, mothers who had were 3.6 times [95% confidence interval (95% CI) 1.05, 12.5] more likely to report elevated anxiety symptoms and 3.6 times [95% CI 1.05, 12.3] more likely to report elevated depressive symptoms when their VPT child was seven years old. Conclusions: The results of this study highlight the need for monitoring and offering ongoing support to bereaved mothers with surviving VPT children. However, within the context of VPT birth, multiple birth does not increase the risk for maternal psychological distress in early childhood. © 2016 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Mothers of infants born very preterm (VPT; b32 weeks' gestation age) report higher rates of psychological distress during infancy and early childhood than mothers of full term infants [for a review, see 1]. Parents of preterm infants (b37 weeks' gestation) are also more likely to report higher rates of parenting stress [2,3], and a more negative influence on the family across early childhood (such as poorer family functioning and financial difficulties) [1,4]. Parents of VPT multiples (twins and triplets) may be at greater risk for psychological and family functioning difficulties given the additional practical, emotional and parenting demands. Twins and higher order multiples ⁎ Corresponding author at: Department of Psychology and Counselling, LaTrobe University, Bundoora, Victoria 3086, Australia. E-mail address:
[email protected] (K. Treyvaud).
http://dx.doi.org/10.1016/j.earlhumdev.2016.04.005 0378-3782/© 2016 Elsevier Ireland Ltd. All rights reserved.
are more likely to be born VPT (11% and 44% respectively) than singletons (1.4% VPT) [5]. Multiple birth is associated with medical, psychosocial, developmental, and economic consequences for families [6]. In mothers of multiples, previous research reported more psychological distress during their child's first year [7,8], at 12 months [9], and at five years [10] compared with mothers of singletons. Mothers of multiples also report higher levels of parenting stress in the preschool years [11,12]. In the newborn period no difference in depressive symptoms between mothers of triplets and mothers of singletons has been reported [9]. One study found lower levels of depression in mothers of preterm multiples compared with mothers of preterm singletons, however the study also reported a slower rate of improvement in depressive symptoms in mothers of preterm multiples from birth to two years [13]. Interviews with parents have suggested that multiple birth also influences the broader family unit, with some parents of multiples reporting increased
2
K. Treyvaud et al. / Early Human Development 100 (2016) 1–5
financial burden, challenges meeting needs of other children, and an impact on the couple's relationship (negative and/or positive) [14]. In general, higher risk infants such as those born VPT have not been the focus of the research relating to multiple births. Furthermore, quantitative research assessing parent and family functioning after VPT multiple birth is limited. Thus, the first aim of the study was to examine the influence of multiple birth (twins or triplets) on mental health, parenting stress and family functioning for mothers of children born VPT at two and seven years' corrected age. The fetal and neonatal death rate in multiple pregnancies is higher than that of singletons [5], and bereaved parents with a surviving twin or triplet may experience a complicated grieving process [15]. Loss of an infant has been identified by parents of multiples as strongly linked with quality of life [14], and compared with mothers of singletons, mothers of children whose co-twin had died were three times more likely to report clinically significant depressive symptoms five years after the birth [10]. The second aim of the current study was to investigate whether bereavement within the context of VPT birth influenced maternal mental health, parenting stress and family functioning. 2. Materials and methods 2.1. Participants Participants were families from the Victorian Infant Brain Studies (VIBeS) cohort, which included 224 infants born at b30 weeks' gestation or with a birthweight b1250 g at the Royal Women's Hospital, Melbourne Australia, between 2001 and 2003. As part of a broader longitudinal observational study assessing child and family outcomes across childhood, follow up questionnaires were completed at two, five and seven years of age (corrected). From the original sample, a complete set of questionnaire data was missing for 30 children, resulting in a sample of 194 children (129 singletons, 65 multiples) from 162 mothers with either all or some questionnaire data. Families were classified as either having living singleton children (singleton group) or living twin or triplet children (multiple group). Families who had a living singleton child from a multifetal pregnancy were classified as singletons for the purpose of the first aim (singletons vs multiples). For the second aim, families were classified into two groups: bereavement (death of an infant from a multifetal pregnancy during either the fetal or neonatal period), and no bereavement. This longitudinal study was approved by the Human Research Ethics Committees of the Royal Women's Hospital and the Royal Children's Hospital, and informed written consent was obtained from parents. 2.2. Measures Maternal mental health was assessed at two years using the total score from the General Health Questionnaire (GHQ) [16]. Higher scores represent increased symptom severity (range 0 to 84), and scores of 24 or greater indicate “clinically significant” symptoms of mental health problems [17]. At seven years, maternal mental health was measured using the Hospital Anxiety and Depression Scale (HADS) [18]. For each of the anxiety and depression domains, scores in the mild to severe range (8–21) were classified as “elevated”, and scores in the moderate to severe range (11–21) were classified as “clinically significant”. There is evidence that the HADS performs well in assessing the severity of anxiety and depression in primary care patients and the general population [19]. The total parent scale from the Parenting Stress Index (PSI) [20] was completed by parents at two and seven years, providing a measure of overall parent-related stress (stress from personal distress, parentchild interaction and child's behavioral characteristics). Higher scores indicate more stress (range 69–188). The PSI has acceptable testretest reliability and construct validity [20]. Family functioning was assessed using the Family Assessment Device (FAD) [21] at both time
points (two and seven years). The general functioning domain was used to provide an overall measure of family functioning (total score range 1–4, where higher scores represent unhealthy family functioning). The FAD has acceptable reliability and concurrent and discriminant validity [21,22]. Responses varied a little between measures, resulting in the following sample sizes: GHQ (n = 143), HADS (n = 130), FAD (n = 140 at two years, 131 at seven years). Outcome data on the PSI was collected per child (rather than per mother), resulting in n = 171 at two years, 154 at seven years. At two and seven years, familial social risk was calculated from a composite measure assessing family structure, education of primary caregiver, occupation and employment of primary income earner, language spoken at home and maternal age when the child was born [23]. Each domain was scored on a three-point scale, where zero represented lowest risk and two represented highest risk, summed to give a total score 0–12. 2.3. Statistical analysis Data were analysed using Stata 13 [24]. Mental health and family functioning were assessed per mother, and parenting stress was assessed for each individual child. To examine whether multiple birth and bereavement were associated with maternal mental health and family functioning, separate linear and logistic regression models were fitted to each continuous and categorical outcome respectively with an indicator for group (multiple vs singleton; no bereavement vs bereavement). To examine whether multiple birth and bereavement were associated with greater parenting stress, linear regression models were fitted at the child level using Generalised Estimating Equations (GEEs) with an exchangeable correlation structure and robust standard errors to allow for correlations between multiples [25]. Analyses were repeated adjusting for concurrent social risk. The overall patterns and magnitude of associations were interpreted, rather than solely focusing on p-values to judge statistical significance [26]. 3. Results Compared with mothers who completed questionnaires, those who did not were more likely to be older at the infant's birth (M(SD) = 37.0 (1.9) years, vs M(SD) = 32.0 (0.4) years, p = 0.003), and have higher social risk scores when their children were age two (M(SD) = 4.27 (1.15) vs M(SD) = 2.36 (0.15), p = 0.006) and seven years (M(SD) = 3.40 (1.06), vs M(SD) = 2.10 (0.13), p = 0.03). At age two, compared with mothers of singletons, mothers of multiples (mothers of twins, n = 27; mothers of triplets, n = 5) were more highly educated, were in families with a more highly skilled primary income earner, and were more likely to be in an intact family (e.g., no shared or single custody), and therefore had lower social risk scores at two overall (Table 1). 3.1. Multiple birth and maternal and family outcomes There was little evidence that maternal mental health at two years, or parenting stress and family functioning at two or seven years was different for mothers of singletons vs multiples (Tables 2 and 3). There was some evidence that mothers of multiples reported lower levels of depressive symptoms at 7 years, but this relationship did not persist after adjusting for concurrent social risk (β [95% CI] = −1.31 [−2.93, 0.30], p = 0.11). There was less evidence of an association between multiple birth and anxiety. The singleton and multiple groups did not differ in maternal mental health symptom classification (Table 3). Although non-significant, there was a pattern for fewer mothers with multiples to be classified in the ‘elevated’, or ‘clinically significant’ categories for depression at seven years. Adjustment for social risk did not alter this pattern of results.
K. Treyvaud et al. / Early Human Development 100 (2016) 1–5 Table 1 Descriptive characteristics of mothers with at least 1 data point at two or seven years. Singletons Multiples (n = 129) (n = 33) Primipara, n (%) Maternal age at birth, M (SD) Maternal education at two, %: (tertiary; secondary school; bsecondary school) Maternal education at seven, %: (tertiary; secondary school; bsecondary school) Occupation of primary income earner at two, %: (professional; semi-skilled; unskilled) Occupation of primary income earner at seven, %: (professional; semi-skilled; unskilled) Family structure at two, %; (intact; dual custody; single parent) Family structure at seven, %; (intact; dual custody; single parent) Social risk at 2, median (IQR) Social risk at 7, median (IQR)
85 (66%) 30 (6) 22%; 64%; 14% 38%; 52%; 9% 34%; 32%; 34% 36%; 55%;9% 85%; 3%; 12% 74%; 14%; 13% 1 (2, 4) 2 (1, 3)
24 (75%) 30 (6) 45%; 15%; 1%⁎ 57%; 33%; 10% 58%; 33%; 13%⁎ 57%; 37%; 6% 97%; 0%; 3%⁎ 80%; 13%; 7% 1 (0, 2)⁎⁎⁎ 2 (0, 2)⁎
Note. M = mean, SD = standard deviation, IQR = interquartile range. ⁎ p b 0.05. ⁎⁎ p b 0.01. ⁎⁎⁎ p b 0.001.
3.2. Bereavement and maternal and family outcomes There were 15 mothers (9%) with data for at least one outcome who experienced bereavement associated with a multifetal pregnancy (14 mothers with one surviving child; one mother with two survivors from a triplet pregnancy). While there was no association between bereavement and mental health difficulties at two years, bereaved mothers reported more anxiety symptoms at seven (Table 4), which persisted after controlling for concurrent social risk (β [95% CI] = 2.60 [0.25, 4.95], p = 0.03). There was little evidence that bereavement influenced parenting stress and family functioning at two or seven years (Table 4). Adjusting for social risk did not change these results. Bereaved mothers were also more likely to report ‘elevated’ symptoms of anxiety and depression (versus scoring in the normal range) at seven years compared with mothers who had not experienced
3
bereavement (Table 5). These results remained significant after controlling for concurrent social risk (anxiety: OR [95% CI] = 4.12 [1.17, 14.48], p = 0.03; depression: OR [95% CI] = 4.67 [1.29, 16.93], p = 0.02). 4. Discussion This study provided evidence that mental health, stress and family functioning of mothers of VPT multiples are similar to mothers of VPT singletons across early childhood. Indeed, there was a pattern in the results for mothers of VPT multiples to report marginally lower levels of depressive symptoms at seven years compared with mothers of VPT singletons. These results are in contrast to those reporting a negative influence of multiple births on maternal depression [7,10] and parenting stress [11,12], but consistent with others showing similar (if not slightly better in the context of preterm birth) adjustment for mothers of multiples in the newborn period [9,13]. The results from the current study may be influenced by the characteristics of the mothers of multiples, who were more socially advantaged and educated than mothers of singletons, a pattern consistent with several previous studies [7,14]. Social risk factors have been linked with poorer mental health for parents of preterm infants [1,13]. One could speculate that the lower social risk status of the multiple group in the current study helped to buffer these mothers from poorer outcomes, and we found some evidence to support this, as the pattern for fewer depressive symptoms in mothers of multiples at seven years weakened when social risk was accounted for. While others have found that the relationship between multiples and increased depressive symptoms persisted even when social risk was accounted for [7,10], overall we found that controlling for social risk did not change the pattern of our results suggesting similar outcomes after VPT birth for families with singletons and multiples. Results from this study indicated that within the context of VPT birth, bereaved mothers were 3.6 times more likely to report elevated anxiety and depressive symptoms when their child was seven years old. The 12 month prevalence rates of anxiety and affective disorders for Australian women are 18% and 7% respectively [27], thus our high rates of anxiety (69% elevated, 23% clinically significant range) and depressive symptoms (38% elevated, 8% clinically significant) seven
Table 2 Maternal mental health, parenting stress and family functioning at two and seven years predicted by group (singleton, multiples). Two years
Mental health
Parenting stress Family functioning
Seven years
Multiples
Singletons
M (SD) n
M (SD) n
16.89 (9.94) 27
115.10 (31.43) 52 1.71 (0.48) 28
18.80 (10.46) 116
108.21 (26.44) 119 1.73 (0.46) 112
β [95% CI]
−1.90 (−6.28, 2.48)
6.88 (−6.08, 19.85) −0.02 (−0.21, 0.17)
p
0.39
0.30 0.83
Multiples
Singletons
M (SD) n
M (SD) n
Anxiety: 5.75 (3.71), 26 Depression:2.56 (2.53) 26 114.02 (27.29) 53 1.67 (0.46) 29
Anxiety: 7.26 (4.22) 102 Depression:4.25 (3.96) 102 119.36 (29.08) 101 1.59 (0.47), 102
β [95% CI]
p
−1.51 (−3.30, 0.28)
0.09
−1.69 (−3.30, −0.07)
0.04
−5.43 (−17.15, 6.27) 0.08 (−0.12, 0.27)
0.36 0.44
Table 3 Maternal mental health (elevated and clinically significant symptoms) at two and seven years predicted by group (singletons, multiples). Elevated symptoms
Mental health at two Anxiety at seven Depression at seven
Multiples
Singletons
n (%)
n (%)
Clinically significant symptoms OR (95% CI)
p
b
8 (31%) 2 (8%)
45 (44%) 20 (20%)
0.56 (0.22, 1.41) 0.34 (0.07, 1.57)
0.22 0.17
Multiples
Singletons
am
n (%)
n (%)
OR (95% CI)
6 (21%) 3 (12%) 0 (0%)
31 (27%) 23 (23%) 7 (7%)
0.73 (0.27, 1.97) 0.44 (0.12, 1.62) Chi2 = 1.88a
Note. Elevated = mild, moderate or severe symptoms; clinically significant = moderate or severe symptoms; % is based on n = available data. a Chi2 test used due to 0 clinically significant in the multiples group. b There is no “elevated” classification on the General Health Questionnaire.
p
0.54 0.22 0.17
4
K. Treyvaud et al. / Early Human Development 100 (2016) 1–5
Table 4 Maternal mental health, parenting stress and family functioning at two and seven years predicted by bereavement. Two years
Mental health
Parenting stress Family functioning
Seven years
Bereavement
No bereavement
M (SD) n
M (SD) n
20.31 (10.22) 13
18.24 (10.39) 130
120.4 (23.2, 14) 1.78 (0.56, 13)
109.4 (28.4, 157) 1.72 (0.45, 127)
β [95% CI]
2.06 (−3.91, 8.03)
11.03 (−2.01, 24.07) 0.06 (−0.20, 0.33)
p
0.50
0.10 0.64
Bereavement
No bereavement
M (SD) n
M (SD, n)
Anxiety: 9.12 (3.70) 13 Depression: 5.42 (3.81) 13 125.2 (26.7, 14) 1.63 (0.39, 13)
Anxiety: 6.71 (4.14) 115 Depression: 3.73 (3.74) 115 116.8 (28.6, 140) 1.60 (0.48, 118)
β [95% CI]
p
2.40 (0.02, 4.77)
0.05
1.69 (−0.48, 3.86)
0.12
9.17 (−6.00, 24.34) 0.03 (−0.24, 0.30)
0.24 0.81
Table 5 Maternal mental health (elevated and clinically significant symptoms) at two and seven years predicted by bereavement. Elevated symptoms Bereavement n (%) Mental health at two Anxiety at seven Depression at seven
9 (69%) 5 (38%)
No bereavement
Clinically significant symptoms p
Bereavement
n (%)
n (%)
n (%)
a
0.04 0.04
4 (31%) 3 (23%) 1 (8%)
33 (26%) 23 (20%) 6 (5%)
44 (38%) 17 (15%)
OR (95% CI)
3.63 (1.05, 12.5) 3.60 (1.05, 12.33)
No bereavement
OR (95% CI)
p
1.29 (0.37, 4.48) 1.2 (0.31, 4.72) 1.51 (0.17, 13.65)
0.69 0.79 0.71
Note. Elevated = mild, moderate or severe symptoms; clinically significant = moderate or severe symptoms; % is based on n = available data. a There is no “elevated” classification on the General Health Questionnaire.
years after the bereavement occurred are clinically important. Others have reported similarly high rates of distress following bereavement, with one study finding that 53% of mothers of five year old children whose co-twin had died reported clinically significant symptoms of depression [10]. In contrast, there was little evidence for group (bereaved, non-bereaved) differences on maternal mental health two years after the birth. It is possible that the effect of bereavement fluctuates over time or may be delayed, perhaps because having a surviving VPT child to care for in the early years influences the grieving process for parents. Although we don't know the content of the elevated maternal anxiety and depressive symptoms, it is possible that the death of a child from a multifetal pregnancy increases anxiety about outcomes of the surviving child or children, who may also serve as a reminder of the infant who died. This may also be confounded by additional stress associated with the increased risk for developmental, academic and emotional challenges for VPT children [29]. There was little evidence for group differences on parenting stress and family functioning at either two or seven years, which suggests that the effects of bereavement after VPT birth may be specific (on the mental health of mothers), rather than general (on the family). This study is the first to our knowledge to examine the influence of multiple birth on maternal and family outcomes within an entirely VPT sample, thereby removing the potentially confounding predictive effect of VPT birth on these outcomes. A further strength of the current study is the inclusion of the multiple assessments of maternal and family outcomes, which continues well after the perinatal period. Limitations of the study include missing maternal data, and mothers with missing data being older and from higher social risk backgrounds. However, the missing data were from both mothers of singletons and multiples, suggesting that any possible effect would not be systematic. Another possible limitation is that our sample was recruited from one hospital where well developed pastoral care is provided for families of VPT infants, multiples, or where there is fetal or neonatal death, which might minimise symptoms of anxiety or depression. Previous research with families of multiples (not exclusively VPT populations) has documented the impact of multiples on the marriage, children's and family needs, parental social isolation and financial burden [14,30]. Research has also demonstrated increased rates of mental health and family functioning difficulties after VPT birth compared with families with full term infants [1]. However, the results from the current study suggest that within the context of VPT birth, having
multiples (twins or triplets) does not increase the risk for difficulties with maternal mental health, parenting stress or family functioning at two or seven years corrected age. Importantly though, mothers of VPT children who experienced bereavement from a multifetal pregnancy were more likely to report elevated anxiety and depressive symptoms when their surviving child/children were seven years old compared with non-bereaved mothers. Clinical implications from this study include the need for bereaved mothers with surviving VPT infants to be monitored and offered support post-discharge, but also in the longer term, as initial assessments of maternal coping and mental health in the early years may not reveal any difficulties. Financial disclosure and conflict of interest The authors have no financial relationships or conflicts of interest relevant to this article to disclose. All authors had significant input into the design and analysis of the study, as well as reviewing and revising the manuscript. Funding support This work was supported by the National Health and Medical Research Council (Project Grants 237117, 491209; Senior Research Fellowship 1081288 to P.J.A); National Institutes of Health (NIH) (R01 HD05709801); The Royal Women's Hospital Research Foundation; the Brockhoff Foundation; and the Murdoch Childrens Research Institute and the Victorian Government's Operational Infrastructure Support Program.
Acknowledgements We would like to acknowledge the contributions of the Victorian Infant Brain Studies team, and thank the families involved in the research. References [1] K. Treyvaud, Parent and family outcomes following very preterm or very low birth weight birth: a review, Semin. Fetal Neonatal Med. 19 (2014) 131–135. [2] R. Schappin, L. Wijnroks, M.M.A.T. Uniken Venema, M.J. Jongmans, Rethinking stress in parents of preterm infants: a meta-analysis, PLoS One 8 (2013), e54992.
K. Treyvaud et al. / Early Human Development 100 (2016) 1–5 [3] K. Treyvaud, K.J. Lee, L.W. Doyle, P.J. Anderson, very preterm birth influences parental mental health and family outcomes seven years after birth, J. Pediatr. 164 (2014) 515–521. [4] C.M.G. Cronin, C.R. Shapiro, O.G. Casiro, M.S. Cheang, The impact of very low birthweight infants on the family is long lasting, Arch. Pediatr. Adolesc. Med. 149 (1995) 151e8. [5] Li Z, Zeki R, Hilder L & Sullivan EA. Australia's mothers and babies 2010. Perinatal Statistics Series No. 27. Cat. No. PER 57. 2012; Canberra: AIHW National Perinatal Epidemiology and Statistics Unit. [6] L.G. Leonard, J. Denton, Preparation for parenting multiple birth children, Early Hum. Dev. 82 (2006) 371–378. [7] Y. Choi, D. Bishai, C.S. Minkovitz, Multiple births are a risk factor for postpartum maternal depressive symptoms, Pediatrics 123 (2009) 1147–1154. [8] L.E. Ross, K. McQueen, S. Vigod, C.L. Dennis, Risk for postpartum depression associated with assisted reproductive technologies and multiple births: a systematic review, Hum. Reprod. Update 17 (2011) 96–106. [9] R. Feldman, A.I. Eidelman, Parent-infant synchrony and the social-emotional development of triplets, Dev. Psychol. 40 (2004) 1133–1147. [10] K. Thorpe, J. Golding, I. MacGillivray, R. Greenwood, Comparison of prevalence of depression in mothers of twins and mothers of singletons, Br. Med. J. 302 (1991) 875–878. [11] K.F. Lutz, C. Burnson, A. Hane, A. Samuelson, S. Maleck, J. Poehlmann, Parenting stress, social support, and mother-child interactions in families of multiple and singleton preterm toddlers, Fam. Relat. 61 (2012) 642–656. [12] F. Olivennes, S. Golombok, C. Ramogida, J. Rust, the Follow-Up Team, Behavioral and cognitive development as well as family functioning of twins conceived by assisted reproduction: findings from a large population study, Fertil. Steril. 84 (2005) 725–733. [13] J. Poehlmann, A.J. Miller Schwichtenberg, D. Bolt, J. Dilworth-Bart, Predictors of depressive symptom trajectories in mothers of infants born preterm or low birthweight, J. Fam. Psychol. 23 (2009) 690–704. [14] M.A. Ellison, J.E. Hall, Social stigma and compounded losses: quality-of-life issues for multiple-birth families, Fertil. Steril. 80 (2003) 405–414. [15] E. Bryan, The impact of multiple preterm births on the family, Br. J. Obstet. Gynaecol. 110 (2003) 24–28.
5
[16] D. Goldberg, P. Williams, A User's Guide to the General Health Questionnaire (GHQ), The NFER-NELSON Publishing Company, Windsor, Berkshire, 1988. [17] D.P. Goldberg, R. Gater, N. Sartorius, T.B. Ustun, O.G. Piccinelli, C. Rutter, The validity of two versions of the GHQ in the WHO study of mental illness in general health care, Psychol. Med. 27 (1997) 191–197. [18] A.S. Zigmond, R.P. Snaith, The hospital anxiety depression scale, Acta Psychiatr. Scand. 67 (1983) 361–370. [19] I. Bjelland, A.A. Dahl, T.T. Haug, D. Neckelmann, The validity of the hospital anxiety and depression scale. An updated literature review, J. Psychosom. Res. 52 (2002) 69–77. [20] R.R. Abidin, Parenting Stress Index, 3rd ed. Odessa, Florida, Psychological Assessment Resources, 1995. [21] N.B. Epstein, L.M. Baldwin, D.S. Bishop, The McMaster family assessment device, J. Marital. Fam. Ther. 9 (1983) 171–180. [22] I.W. Miller, N.B. Epstein, D.S. Bishop, G.I. Keitner, The McMaster family assessment device: reliability and validity, J. Marital. Fam. Ther. 11 (1985) 345–356. [23] G. Roberts, K. Howard, A.J. Spittle, N.C. Brown, P.J. Anderson, L.W. Doyle, Rates of early intervention services in very preterm children with developmental disabilities at age two, J. Paediatr. Child Health 44 (2008) 276–280. [24] StataCorp, Stata Statistical Software: Release 13, StataCorp LP, College Station, TX, 2013. [25] J.B. Carlin, L.C. Gurrin, J.A.C. Sterne, R. Morley, T. Dwyer, Regression models for twin studies: a critical review, Int. J. Epidemiol. 34 (2005) 1089–1099. [26] J.A.C. Stern, G.D. Smith, Sifting the evidence — what's wrong with significance tests? Br. Med. J. 322 (2001) 226–231. [27] Australian Bureau of Statistics, National Survey of Mental Health and Wellbeing: Summary of Results. ABS Catalogue No. 4326.0, Australian Bureau of Statistics, Canberra, 2008. [29] P. Anderson, L.W. Doyle, Neurobehavioral outcomes of school-age children born extremely low birth weight or very preterm in the 1990s, J. Am. Med. Assoc. 289 (2003) 3264–3272. [30] J. Fisher, A. Stocky, Maternal perinatal mental health and multiple births: implications for practice, Twin Res. 6 (2003) 506–513.