Acta Pædiatrica ISSN 0803-5253
REGULAR ARTICLE
Health-related quality of life may deteriorate from adolescence to young adulthood after extremely preterm birth Berit B atsvik ([email protected])1, Bente J. Vederhus2,3,4, Thomas Halvorsen2,3, Tore Wentzel-Larsen5,6,7, Marit Graue2,8, Trond Markestad2,3 1.Betanien University College, Bergen, Norway 2.Department of Pediatrics, Haukeland University Hospital, Bergen, Norway 3.Department of Clinical Science, University of Bergen, Bergen, Norway 4.Department of Global Public Health and Primary Care, University of Bergen, Bergen, Norway 5.Centre for Clinical Research, Haukeland University Hospital, Bergen, Norway 6.Centre for Child and Adolescent Mental Health, Eastern and Southern Norway, Oslo, Norway 7.Norwegian Centre for Violence and Traumatic Stress Studies, Oslo, Norway 8.Centre for Evidence-Based Practice, Bergen University College, Bergen, Norway
Keywords Adolescence, Extremely premature infant, Health-related quality of life, Self-rated health, Young adulthood Correspondence Berit B atsvik, Betanien University College, Vestlundveien 19, N-5145 Fyllingsdalen, Norway. Tel: +47 55507278 | Fax: +47 55507301 | Email: [email protected] Received 11 February 2015; revised 23 April 2015; accepted 3 June 2015. DOI:10.1111/apa.13069
ABSTRACT Aim: This study examined the development of health-related quality of life (HRQoL) and health from adolescence to adulthood after extremely preterm birth. Methods: We assessed a population-based cohort of extremely preterm-born (EPB) infants (gestational age of ≤28 weeks or birthweight of ≤1000 grams) and term-born (TB) controls at 17 and 24 years of age. They completed the Child Health Questionnaire-Child Form 87 at 17 years of age, the Short Form Health Survey-36 (SF-36) at 24 years of age and the Health Behaviour in School-aged Children–Symptom Checklist at both ages. Results: Of the 51 eligible EPB subjects, 46 (90%) were included and nine had severe neurosensory disabilities. On the whole, EPB and TB subjects gave their HRQoL and health similar ratings, but EPB subjects with disabilities reported poorer physical functioning at 17 and EPB subjects without disabilities reported lower scores on three of the eight SF-36 scales for social functioning and mental health and reported more psychological health complaints at 24. Differences remained in adjusted analyses. Changes from 17 to 24 years of age were minor in EPB subjects with disabilities. Conclusion: Our comparison of EPB and TB subjects at the ages of 17 and 24 indicated that psychosocial HRQoL may deteriorate for EPB subjects when they enter adulthood.
BACKGROUND Extremely preterm-born (EPB) children tend to have lower cognitive, motor and behavioural abilities than term-born (TB) children during childhood (1,2). Nevertheless, they have generally reported that they cope well with school (1) and have tended to have a positive view on the future, both as children and adolescents (3,4). They have also considered their health-related quality of life (HRQoL) and abilities as good as their term-born peers (5,6). However, parents have assessed their preterm-born children’s HRQoL as being poorer (7) and it has been discussed whether the positive self-reports of HRQoL were reliable, demonstrated a lack of realistic insight or reflected adjustment and a kind of resilience (4,8).
In general, young adults born preterm have not rated their HRQoL significantly differently from normal birthweight peers, but there have been few studies limited to EPB subjects, both for adolescents (3,9) and young adults (10–12). We are not aware of any studies on how perception of different aspects of HRQoL may change from adolescence to adulthood.
Key notes
Abbreviations CHQ-CF 87, Child Health Questionnaire-Child Form 87; EPB, Extremely preterm-born; HBSC-SCL, Health Behaviour in School-aged Children–Symptom Checklist; HRQoL, Healthrelated quality of life; SF-36, Short Form Health Survey-36; TB, Term-born.
948
We examined the development of health-related quality of life (HRQoL) and health from adolescence to adulthood after extremely preterm birth. In general, extremely preterm-born (EPB) and termborn subjects gave their HRQoL and health similar ratings at 17 years of age, but EPB subjects with disabilities reported poorer physical functioning. At 24 years of age, EPB subjects without disabilities reported lower scores for social functioning and mental health and reported more psychological health complaints.
©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2015 104, pp. 948–955
B atsvik et al.
Entering adulthood means new and demanding challenges, such as coping with higher education and obtaining satisfying jobs and intimate relationships. Moster et al. (13) reported that adults born extremely preterm had less success in many of these areas, and our hypothesis was that the transition to the demands of adulthood might have a negative effect on perception of life. Therefore, the aim of this study was to compare the development of self-perceived HRQoL and health complaints from adolescence to young adulthood in EPB subjects and matched TB controls.
PATIENTS AND METHODS Participants All EPB survivors born to mothers living within a defined area of the Western Norway Regional Health Authority from January 1, 1982 to December 31, 1985, were invited to participate in the study in 2001–2002 when they were 15– 19 years of age (mean 17.6 years). We defined EPB was as a gestational age ≤28 weeks or birthweight of ≤1000 grams. The children were treated at the only neonatal intensive care unit in the region, Haukeland University Hospital. The controls for each preterm child were the next child born of the same gender with a birthweight of between three and four kilograms (Norwegian 10th and 90th percentiles), identified through the birth protocols. If one potential control subject declined, the next-born subject fulfilling the entry criteria was approached. All the subjects were invited for a second follow-up seven years later in 2008–2009. Socio-demographic and clinical data Information on socio-demographic characteristics, such as current educational level, living arrangements and occupation or employment, was obtained from a self-developed composite questionnaire, which largely used questions derived from Norwegian population studies (14). At the first follow-up at 17 years of age, parents reported on their children’s learning difficulties and their previous, or present, needs for educational support at school. They also provided details on their own educational level. The medical history was obtained from parental reports (age 17 years), self-reports (age 24 years), hospital records and by the responsible paediatrician (TH). Comorbidities were common and disability was defined as not being able to live independent lives due to disabling cerebral palsy (CP), deafness or severe hearing loss, blindness or severe vision impairment or severe psychiatric disease (Table 1). In accordance with this, the EPB group was divided into EPB subjects who were healthy or had severe disability. Self-reported data The Child Health Questionnaire-Child Form 87 (CHQ-CF 87) (15) was used to assess HRQoL at the first follow-up at the mean age of 17 years, and the first version of the Short Form Health Survey-36 (SF-36) (16) was used at the second follow-up at the mean age of 24 years. Different questionnaires were used because they were age appropriate and based on The World Health Organization’s definition of
Health-related quality of life after extremely preterm birth
health. Both questionnaires are widely used and therefore allow comparison with other studies. Child Health Questionnaire-Child Form 87 (CHQ-CF 87) This 87-item generic instrument is designed to measure functional health and well-being of children and adolescents, as perceived by the youths themselves. The CHQCF87 assesses self-reported physical, emotional and social well-being on nine scales: the physical functioning scale (nine items describing ability to do physical tasks), the rolephysical scale (three items assessing limitations in schoolrelated activities and activities with friends caused by problems in physical health), the bodily pain scale (two items assessing the intensity and frequency of general pain and discomfort), the general health scale (12 items assessing the perception of overall health), the behaviour scale (17 items assessing aggression, delinquency, hyperactivity/ impulsivity and social withdrawal), the self-esteem scale (14 items assessing satisfaction with school and athletic ability, looks and appearance, ability to get along and overall feelings about life), the mental health scale (16 items assessing anxiety, depression and positive aspects), the roleemotional scale (three items assessing limitations in activities with friends and at school caused by emotional problems) and the role-behavioural scale (three items assessing limitations in activities with friends and at school caused by behavioural problems). The responses are rated along four to six ordinal response levels of agreement to a certain categorical statement, such as ‘very often’ to ‘not at all’. The preceding four weeks are used as the recall period, except for general health, which pertains to present status. The raw score for each CHQ-CF87 scale was based on the mean of valid items if at least half of the item in each scale was valid and transformed into a scale of zero (poor) to 100 (optimal), with higher scores indicating better functional health and well-being. A Norwegian version of this instrument with acceptable psychometric properties was used (17). Short Form Health Survey-36 (SF-36, first version) This instrument consists of 36 items across eight domains addressing physical and mental health. The domains are physical functioning (10 items describing ability to do physical tasks), role-physical (four items concerning difficulties or limitations in physical ability), bodily pain (two items), general health (five items), vitality (four items measuring level of energy), social functioning (two items about physical and mental health interfering with normal social life), role-emotional (three items about effects of mental problems on everyday life) and mental health (five items measuring moods). Except for the two role-functioning scales, which have dichotomised response choices, the responses are rated along three to six-point Likert-type scales with the preceding four weeks as the recall period, except for physical functioning and general health, which pertain to current status. The raw score for each SF-36 scale was based on the mean of valid items if at least half of the item in each scale was valid, and then transformed into a
©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2015 104, pp. 948–955
949
B atsvik et al.
Health-related quality of life after extremely preterm birth
Table 1 Demographic and clinical characteristics at 17 and 24-years-of-age in subjects born extremely preterm (EPB) with or without severe disability, and at term (TB)
Birth characteristics Gender, male/female, n Gestational age, weeks, mean (SD) Birth weight, gram, mean (SD) Age first follow-up, years, mean (SD) Age second follow-up, years, mean (SD) Neurosensory impairments, n Disabling CP Wheelchair Non-disabling CP Deaf/hard of hearing Blind/visually impaired Parents’ education Mother’s highest education, n (%) Primary school Upper secondary College/University Father’s highest education, n (%) Primary school Upper secondary College/University School and school support at 17 years, n (%)‡ Mainstream school School psychological counseling Academic assistance Education and employment at 24 years, n (%) Highest educational attainment Primary/secondary school College/University, four years Employment Still in education Working Redundancy Disability pension Living arrangements at 17 years/24 years, n Parents’ home§ Sheltered housing/institution Independent living With spouse or partner
EPB severe disability (n = 9)*
EPB healthy (n = 37)**
TB controls (n = 46)***
6/3 27.0 (1.9) 946 (176) 17.9 (1.7) 24.4 (1.8)
19/18 27.4 1029 17.6 24.2
25/21 (1.3) (194) (1.1) (1.2)
EPB severe disability vs TB controls p-value†
EPB healthy vs TB controls p-value†
3441 (310) 17.8 (1.2) 24.6 (1.2)
4 3 3 2 7
0 0 4 1 2
0 0 0 0 0
3 (43) 2 (29) 2 (29)
7 (24) 12 (41) 10 (35)
4 (11) 16 (43) 17 (46)
0.319
0.302
3 (43) 3 (43) 1 (14)
9 (30) 11 (37) 10 (33)
5 (14) 14 (39) 17 (47)
0.419
0.272
7 (78) 7 (78) 6 (67)
35 (95) 23 (62) 18 (49)
42 (91) 16 (35) 8 (17)
0.333 0.017 0.004
0.251 0.012 0.002
7 (88) 1 (12) 0
17 (49) 14 (40) 4 (11)
14 (35) 14 (35) 12 (30)
0.021
0.032
2 (25) 1 (12)
7 (20) 24 (69)
0.003
0.042
5 (63)
4 (11)
14 (35) 25 (63) 1 (3) 0
0.006/0.104
0.808/0.111
6/0 3/1 0/6 0/1
34/6 0/1 2/13 1/15
43/4 0/0 3/17 0/19
All information about demographic and clinical characteristics were obtained from a general questionnaire. *8 at second follow-up, **35 at second follow-up, ***40 at second follow-up. †Pearson’s chi square test. ‡Questions about learning difficulties were reported by parents. § With one parent or both. The bold values are statistically significant.
scale from 0 to 100, with higher scores indicating better functional health and well-being. A validated Norwegian version with acceptable reliability and validity was used (18). Health Behaviour in School-aged Children–Symptom Check List The HBSC-SCL questionnaire that was used at both measurement points assesses frequency of subjective health complaints during the previous six months. It is an eight-item scale addressing physical complaints, such as
950
headache, abdominal pain, back pain and dizziness (somatic subscore), and psychological complaints, such as feeling low, nervous, irritable and difficulty falling asleep (psychological subscore). Ratings were made on a five-point frequency scale with ‘rarely or never’ as zero and ‘approximately every day’ as four. An overall sum score of 0 to 32 and two subscores of 0–16 were generated, and lower scores indicated better subjective physical or psychological health. In this study, the subscores were used. Adequate reliability and validity have been reported in a Norwegian population of adolescents (19).
©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2015 104, pp. 948–955
B atsvik et al.
Health-related quality of life after extremely preterm birth
Ethics The Regional Committee on Medical Research Ethics of the Western Norway Health Region Authority approved the study. Informed written consent was obtained from each participant and from parents at the first follow-up if the subjects were minors. Statistical analyses For comparison of demographic data on a categorical level, differences between groups were analysed using exact chisquared tests. For the CHQ-87, SF-36 and the HBSC-SCL scales, one-way analyses of variance (ANOVA) using Tamhane corrections for multiple testing in post hoc analyses were applied for comparison between the three groups. Results are presented as group means (SD) and mean group differences with 95% confidence intervals (CI). Linear mixed effects models were used to compare development on the HBSC-SCL scores from 17 to 24 years of age between the three groups, including a time by group interaction in unadjusted and adjusted analyses for gender and education at the age of 24 years. p ≤ 0.05 was considered statistically significant. SPSS version 21 (SPSS Inc, Chicago, IL, USA) was used for ANOVA while the R (The R Foundation for Statistical Computing, Vienna, Austria) package nlme was used for mixed effects models.
RESULTS At the first assessment, 46 of 51 eligible EPB subjects (90%) participated. An average of 1.4 TB adolescents was invited for each EPB subject to recruit the matched control group. At the second follow-up, 43 EPB young adults and 40 TB controls (84% and 87%, respectively) participated (Fig. 1). All were Caucasians and native Norwegians.
The first follow-up at 17 years of age Group characteristics Similar proportions of healthy EPB and TB subjects (94%) and 67% of EPB subjects with severe disabilities were living at home with their parents (Table 1). The remaining EPB subjects with disabilities were living in some kind of sheltered housing. Most subjects were attending mainstream secondary schools. Of the healthy EPB adolescents, 49% had received academic support compared to 17% of the TB controls (p = 0.002), and 62% vs 35% (p = 0.012) had received support from the School Psychological Service. Health-related quality of life (CHQ-CF87) From the ANOVA analyses, EPB adolescents with severe disabilities reported lower scores on most domains, but only physical functioning was significantly lower compared to their TB peers (p = 0.034) (Table 2). The healthy EPB subjects and TB controls did not differ significantly on the physical and psychological domains. Subjective health complaints There were no significant differences between the healthy EPB subjects, the EPB subjects with severe disabilities and the TB controls in self-rated somatic and psychological health complaints, as measured by the HBSC-SCL in the ANOVA analyses (Table 2). The second follow-up at 24 years of age Group characteristics Most of the EPB and TB young adults were living independently or together with spouses or partners (Table 1). The EPB subjects with severe disabilities were mainly living independently of their families, while receiving some kind of care services or domestic help.
Eligible infants born extremely preterm and admitted to NICU* care in 1982 – 85 (n = 81)
Control infants born at term
First follow-up at age 15-19 years: 46/51 (90%)
First follow-up at age 15- 19 years: (n = 46)
Died in NICU or after discharge (n = 30) Declines, n = 5
Declines, n = 3
Declines, n = 3
Second follow-up at age 21- 25years 43/51 (84%)
Second follow-up at age 21 to 25 40/46 (87%)
Figure 1 Recruitment of subjects. *Neonatal intensive care unit.
©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2015 104, pp. 948–955
951
B atsvik et al.
Health-related quality of life after extremely preterm birth
Table 2 Self-reported functional health and well-being at adolescence in subjects born extremely preterm (EPB) with or without severe disability and controls born at term (TB) provided as mean scores (standard deviation) on the Child Health Questionnaire-Child Form 87 (CHQ-CF87) and the Health Behaviour School-aged Children-Symptom Checklist (HBSC-SCL) EPB severe disability (n = 9) CHQ-CF87 domain Physical functioning Role - physical Bodily pain General health Self-esteem Behaviour Role - behavior Mental health Role - emotional HBSC-SCL variables Somatic complaints Psychological complaints
Mean (SD) 52.3 (39.4) 81.5 (32.9) 61.1 (35.5) 63.2 (18.8) 71.2 (14.3) 73.5 (11.0) 96.3 (11.1) 75.0 (16.2) 81.5 (33.8) Mean(SD)min-max 2.4 (2.7) 0–6 3.4 (3.7) 0–12
EPB healthy (n = 37)
TB controls (n = 46)
94.5 95.8 73.8 67.0 70.2 80.1 97.3 74.6 86.5
95.1 96.1 73.0 66.7 70.6 79.6 98.6 75.6 94.0
(9.5) (9.9) (27.2) (17.9) (15.5) (12.5) (7.1) (15.8) (23.7)
3.2 (3.6) 0–13 2.9 (3.3) 0–12
(7.7) (10.8) (25.5) (17.7) (12.3) (11.3) (6.9) (12.6) (13.8)
2.9 (3.1) 0–12 3.0 (2.9) 0–14
EPB severe disability vs TB controls Mean difference (95%CI) 42.8 ( 82.3; 3.3)* 14.7 ( 47.5;18.3) 11.9 ( 47.7;23.8) 3.5 ( 22.7;15.6) 0.6 ( 13.8;15.1) 6.1 ( 17.3;5.1) 2.3 ( 13.4;8.9) 0.6 ( 17.0;15.7) 12.5 ( 46.3;21.4) Mean difference (95%CI) 0.4 ( 3.2;2.4) 0.4 ( 3.3;4.2)
EPB healthy vs TB controls
0.59 0.3 0.7 0.2 0.4 0.5 1.3 0.9 7.5
( ( ( ( ( ( ( ( (
5.3;4.1) 5.9;5.2) 13.5;15.0) 9.3;9.8) 8.1;7.2) 6.0;7.0) 5.0;2.5) 8.8;6.8) 18.3;3.4)
0.4 ( 1.5;2.2) 0.1 ( 1.8;1.6)
Differences between groups were analyzed with ANOVA using Tamhane correction for multiple testing in post hoc analyses. The bold value is statistically significant,*p = 0.034. Mixed effects models were used to assess differences in the HBSC-SCL scores between the three groups.
The TB subjects had achieved significantly higher education, and a higher proportion were still in education compared to the healthy EPB subjects (p = 0.042). Five of the eight EPB subjects with severe disabilities and three of the 37 healthy EPB subjects were receiving a disability pension, but none of the TB controls were. Health-related quality of life (SF-36) In the ANOVA analyses, the healthy EPB subjects scored consistently lower on all SF-36 domains and significantly lower on social functioning, emotional role and mental health than the TB controls (Table 3). The EPB subjects with disabilities tended to score lower than the TB controls on most of the scales, but the differences were not significant. The EPB subjects, as a whole, scored significantly lower than the TB controls on five domains of the SF-36. These five domains were bodily pain (mean differences 10.8, 95% CI 12.3 to 0.2), vitality (mean difference 8.9, 95% CI 16.3 to 1.5), social functioning (mean difference 13.3, 95% CI 23.7 to 2.9), emotional role (mean difference 22.8, 95%CI 39.0 to 6.6) and mental health (mean difference 9.6, 95%CI 16.1 to 3.1). Developmental changes in subjective health complaints Due to missing values in the adjustment variables, the adjusted analyses were based on 81 people compared with 92 people in the unadjusted analyses. In the unadjusted mixed effects analysis regarding the psychological health complaint, there was a significant group by time interaction (p = 0.003). The group differences were nonsignificant at adolescence (p = 0.867), but significant at young adulthood (p = 0.005). Specifically, the healthy EPB group had a higher mean score than the TB controls (2.5, 95% CI 1.0 to 4.0, p = 0.001). There were no
952
significant differences between the EPB subjects with disabilities and the TB controls at either age (p ≥ 0.131). For the healthy EPB subjects, the psychological complaint increased from 17 to 24 years of age (1.9, 95% CI 0.8 to 3.0, p = 0.001) and descriptive information is given in Tables 2 and 3 and Fig. 2. In the adjusted analyses regarding psychological health complaint (data not reported in Tables), the interaction was still significant (p = 0.005). The group differences were insignificant at adolescence (p = 0.883), but significant at young adulthood (p = 0.004). Specifically, the healthy EPB subjects had a higher mean score at young adulthood than the TB controls (2.7, 95% CI 1.1–4.2, p = 0.001). There were no significant differences between the EPB subjects with disabilities and the TB controls when they were adolescents or young adults (p ≥ 0.152). For the healthy EPB subjects, the psychological complaint increased (1.8, 95% CI 0.7–3.0, p = 0.002). In the unadjusted mixed effects analysis regarding the somatic health complaint, there was no significant group by time interaction (p = 0.135). Furthermore, there were no significant group differences at any measurement points (p ≥ 0.275) and no significant time differences in any group (p ≥ 0.169). In the adjusted analysis, there was no significant group by time interaction (p = 0.182). Furthermore, there were no significant group differences at any assessment points (p ≥ 0.271) and no significant time differences in any group (p ≥ 0.191).
DISCUSSION At young adulthood, but not at adolescence, the EPB subjects had more psychological complaints and rated their psychosocial HRQoL as inferior to that of their TB peers. In contrast to the TB controls, psychological complaints
©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2015 104, pp. 948–955
B atsvik et al.
Health-related quality of life after extremely preterm birth
Table 3 Self-reported health-related quality of life at 24-years-of-age in subjects born extremely preterm (EPB) with or without severe disabilities and controls born at term (TB) given as mean (standard deviation) scores on the Short Form Health Survey-36 (SF36) and the Health Behaviour School-aged Children-Symptom Checklist (HBSC-SCL) EPB severe disability (n = 8) SF-36 domain Physical functioning Role – physical Bodily pain General health Vitality Social functioning Role – emotional Mental health HBSC-SCL variables Somatic complaints Psychological complaints
Mean (SD) 56.9 (43.0) 68.8 (37.2) 73.9 (28.1) 65.9 (22.2) 53.1 (14.4) 87.5 (17.7) 83.3 (35.6) 75.5 (12.7) Mean (SD)min-max 3.8 (3.0) 0–9 4.0 (2.1) 1–7
EPB healthy (n = 35)
TB controls (n = 43)
92.1 80.5 68.4 74.1 50.7 74.3 61.4 69.5
94.2 86.9 80.2 77.1 60.0 90.0 88.3 80.2
(15.0) (30.4) (25.0) (25.0) (20.5) (30.2) (42.2) (17.8)
3.8 (3.9) 0–16 4.9 (4.5) 0–16§
(11.3) (25.9) (23.0) (20.0) (13.8) (17.3) (31.6) (12.2)
2.4 (2.2) 0–9 2.2 (2.3) 0–12
EPB severe disability vs TB controls Mean difference (95%CI) 37.4 ( 84.7; 9.9) 18.1 ( 59.1;22.8) 6.3 ( 27.2; 24.6) 11.3 ( 35.8; 13.3) 6.9 ( 22.8; 9.0) 2.5 ( 22.1; 17.1) 5.0 ( 44.3; 34.3) 4.7 ( 18.8; 9.4) Mean difference (95%CI) 1.4 ( 1.9; 4.7) 1.8 ( 0.5; 4.1)
EPB healthy vs TB controls
2.2 6.4 11.8 3.0 9.3 15.7 26.9 10.7
( ( ( ( ( ( ( (
9.7;5.4) 22.5; 9.7) 25.4;1.8) 15.9;9.9) 19.4;0.8) 30.0; 1.5)* 48.3; 5.5)† 19.5; 1.9)‡
1.4 ( 0.4;3.3) 2.5 (1.0;4.0) ¶
Differences between groups were analyzed with ANOVA using Tamhane correction for multiple testing in post hoc analyses. *p = 0.026, †p = 0.009, ‡p = 0.011.Mixed effects models were used to assess differences in the HBSC-SCL scores between 17 to 24-years-of-age, and between the three groups at each time, including a group by time interaction. EPB healthy between 17 to 24-years-of-age §p = 0.002. ¶p = 0.001. The bold values are statistically significant.
Haealth complaints (mean)
6
EP severe disability
EP healthy
Term born
5
*
4
**
3
2
1
0 Somatic complaints 17 yr
Somatic complaints 24 yr
Psychological complaints 17 yr
Psychological complaints 24 yr
Figure 2 Development of self-reported somatic and psychological health complaints from 17 to 24 years of age in the extremely preterm-born (EPB) subjects with or without severe disabilities and the term-born (TB) controls according to mean scores on the HBSC-SCL. Higher scores indicate more complaints. Mixed effects models, including group by time interaction, were used to assess differences from 17 to 24 years of age and differences between the three groups at each age of assessment. *p = 0.002 compared to 17 years of age, **p = 0.001 compared to TB peers at 24 years of age.
increased from adolescence to young adulthood. The negative trajectory was particularly noticeable for the healthy EPB subjects. The major strengths of this study were the high participation rate at both assessments and the recruitment of controls in a way that limited bias. A limitation was the relatively low number of participants, rendering the study at a particular risk of type-two errors, for failing to detect differences that may have been present. This limitation was particularly relevant when performing subgroup comparisons, such as
those involving participants with severe disabilities. A priori power calculation could not be performed as the distribution of the CHQ-CF87-scores in EPB subjects was unknown when planning the study. According to statistical advice, post hoc power calculations were not carried out. Instead, we quantified uncertainty by stating 95% confidence intervals and p-values. SF-36 is considered to be the adult counterpart to the CHQ-CF87 as it assesses similar domains of HRQoL (20), but direct comparison of the developmental trajectories could not be performed as the instruments are not constructed in the same way. Consistent with our results, earlier studies have demonstrated that EPB adolescents tend to rate their HRQoL as equal to that of TB peers despite extensively reported increased occurrence of academic underachievement, behavioural problems and various health problems (8,10). In the EPB adolescents with severe disabilities, physical limitation was not accompanied by significantly more limitations in social roles, or poorer psychological health, suggesting that physical ability was not considered an important aspect of HRQoL. These findings may indicate an optimistic attitude to life, ability to cope with life challenges, resilience or adjustment of expectations or changes in internal standards and values, even for those with severe disabilities. This is in line with some previous reports (21,22). There are, however, worrying reports of psychosocial and educational vulnerabilities among prematurely born subjects in late adolescence that may complicate their transition to adult life (8,23,24) and cause less resilience and increased risk of functional problems as adults (9,25). The main outcome of this study may support such concerns. In line with this, a previous longitudinal study demonstrated a nonsignificant tendency towards decrease in the HRQoL of prematurely born adults (13), and another study reported a small decrease for both EPB and TB young adults (22).
©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2015 104, pp. 948–955
953
B atsvik et al.
Health-related quality of life after extremely preterm birth
Using SF-36, young adults with birthweights of
REGULAR ARTICLE
Health-related quality of life may deteriorate from adolescence to young adulthood after extremely preterm birth Berit B atsvik ([email protected])1, Bente J. Vederhus2,3,4, Thomas Halvorsen2,3, Tore Wentzel-Larsen5,6,7, Marit Graue2,8, Trond Markestad2,3 1.Betanien University College, Bergen, Norway 2.Department of Pediatrics, Haukeland University Hospital, Bergen, Norway 3.Department of Clinical Science, University of Bergen, Bergen, Norway 4.Department of Global Public Health and Primary Care, University of Bergen, Bergen, Norway 5.Centre for Clinical Research, Haukeland University Hospital, Bergen, Norway 6.Centre for Child and Adolescent Mental Health, Eastern and Southern Norway, Oslo, Norway 7.Norwegian Centre for Violence and Traumatic Stress Studies, Oslo, Norway 8.Centre for Evidence-Based Practice, Bergen University College, Bergen, Norway
Keywords Adolescence, Extremely premature infant, Health-related quality of life, Self-rated health, Young adulthood Correspondence Berit B atsvik, Betanien University College, Vestlundveien 19, N-5145 Fyllingsdalen, Norway. Tel: +47 55507278 | Fax: +47 55507301 | Email: [email protected] Received 11 February 2015; revised 23 April 2015; accepted 3 June 2015. DOI:10.1111/apa.13069
ABSTRACT Aim: This study examined the development of health-related quality of life (HRQoL) and health from adolescence to adulthood after extremely preterm birth. Methods: We assessed a population-based cohort of extremely preterm-born (EPB) infants (gestational age of ≤28 weeks or birthweight of ≤1000 grams) and term-born (TB) controls at 17 and 24 years of age. They completed the Child Health Questionnaire-Child Form 87 at 17 years of age, the Short Form Health Survey-36 (SF-36) at 24 years of age and the Health Behaviour in School-aged Children–Symptom Checklist at both ages. Results: Of the 51 eligible EPB subjects, 46 (90%) were included and nine had severe neurosensory disabilities. On the whole, EPB and TB subjects gave their HRQoL and health similar ratings, but EPB subjects with disabilities reported poorer physical functioning at 17 and EPB subjects without disabilities reported lower scores on three of the eight SF-36 scales for social functioning and mental health and reported more psychological health complaints at 24. Differences remained in adjusted analyses. Changes from 17 to 24 years of age were minor in EPB subjects with disabilities. Conclusion: Our comparison of EPB and TB subjects at the ages of 17 and 24 indicated that psychosocial HRQoL may deteriorate for EPB subjects when they enter adulthood.
BACKGROUND Extremely preterm-born (EPB) children tend to have lower cognitive, motor and behavioural abilities than term-born (TB) children during childhood (1,2). Nevertheless, they have generally reported that they cope well with school (1) and have tended to have a positive view on the future, both as children and adolescents (3,4). They have also considered their health-related quality of life (HRQoL) and abilities as good as their term-born peers (5,6). However, parents have assessed their preterm-born children’s HRQoL as being poorer (7) and it has been discussed whether the positive self-reports of HRQoL were reliable, demonstrated a lack of realistic insight or reflected adjustment and a kind of resilience (4,8).
In general, young adults born preterm have not rated their HRQoL significantly differently from normal birthweight peers, but there have been few studies limited to EPB subjects, both for adolescents (3,9) and young adults (10–12). We are not aware of any studies on how perception of different aspects of HRQoL may change from adolescence to adulthood.
Key notes
Abbreviations CHQ-CF 87, Child Health Questionnaire-Child Form 87; EPB, Extremely preterm-born; HBSC-SCL, Health Behaviour in School-aged Children–Symptom Checklist; HRQoL, Healthrelated quality of life; SF-36, Short Form Health Survey-36; TB, Term-born.
948
We examined the development of health-related quality of life (HRQoL) and health from adolescence to adulthood after extremely preterm birth. In general, extremely preterm-born (EPB) and termborn subjects gave their HRQoL and health similar ratings at 17 years of age, but EPB subjects with disabilities reported poorer physical functioning. At 24 years of age, EPB subjects without disabilities reported lower scores for social functioning and mental health and reported more psychological health complaints.
©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2015 104, pp. 948–955
B atsvik et al.
Entering adulthood means new and demanding challenges, such as coping with higher education and obtaining satisfying jobs and intimate relationships. Moster et al. (13) reported that adults born extremely preterm had less success in many of these areas, and our hypothesis was that the transition to the demands of adulthood might have a negative effect on perception of life. Therefore, the aim of this study was to compare the development of self-perceived HRQoL and health complaints from adolescence to young adulthood in EPB subjects and matched TB controls.
PATIENTS AND METHODS Participants All EPB survivors born to mothers living within a defined area of the Western Norway Regional Health Authority from January 1, 1982 to December 31, 1985, were invited to participate in the study in 2001–2002 when they were 15– 19 years of age (mean 17.6 years). We defined EPB was as a gestational age ≤28 weeks or birthweight of ≤1000 grams. The children were treated at the only neonatal intensive care unit in the region, Haukeland University Hospital. The controls for each preterm child were the next child born of the same gender with a birthweight of between three and four kilograms (Norwegian 10th and 90th percentiles), identified through the birth protocols. If one potential control subject declined, the next-born subject fulfilling the entry criteria was approached. All the subjects were invited for a second follow-up seven years later in 2008–2009. Socio-demographic and clinical data Information on socio-demographic characteristics, such as current educational level, living arrangements and occupation or employment, was obtained from a self-developed composite questionnaire, which largely used questions derived from Norwegian population studies (14). At the first follow-up at 17 years of age, parents reported on their children’s learning difficulties and their previous, or present, needs for educational support at school. They also provided details on their own educational level. The medical history was obtained from parental reports (age 17 years), self-reports (age 24 years), hospital records and by the responsible paediatrician (TH). Comorbidities were common and disability was defined as not being able to live independent lives due to disabling cerebral palsy (CP), deafness or severe hearing loss, blindness or severe vision impairment or severe psychiatric disease (Table 1). In accordance with this, the EPB group was divided into EPB subjects who were healthy or had severe disability. Self-reported data The Child Health Questionnaire-Child Form 87 (CHQ-CF 87) (15) was used to assess HRQoL at the first follow-up at the mean age of 17 years, and the first version of the Short Form Health Survey-36 (SF-36) (16) was used at the second follow-up at the mean age of 24 years. Different questionnaires were used because they were age appropriate and based on The World Health Organization’s definition of
Health-related quality of life after extremely preterm birth
health. Both questionnaires are widely used and therefore allow comparison with other studies. Child Health Questionnaire-Child Form 87 (CHQ-CF 87) This 87-item generic instrument is designed to measure functional health and well-being of children and adolescents, as perceived by the youths themselves. The CHQCF87 assesses self-reported physical, emotional and social well-being on nine scales: the physical functioning scale (nine items describing ability to do physical tasks), the rolephysical scale (three items assessing limitations in schoolrelated activities and activities with friends caused by problems in physical health), the bodily pain scale (two items assessing the intensity and frequency of general pain and discomfort), the general health scale (12 items assessing the perception of overall health), the behaviour scale (17 items assessing aggression, delinquency, hyperactivity/ impulsivity and social withdrawal), the self-esteem scale (14 items assessing satisfaction with school and athletic ability, looks and appearance, ability to get along and overall feelings about life), the mental health scale (16 items assessing anxiety, depression and positive aspects), the roleemotional scale (three items assessing limitations in activities with friends and at school caused by emotional problems) and the role-behavioural scale (three items assessing limitations in activities with friends and at school caused by behavioural problems). The responses are rated along four to six ordinal response levels of agreement to a certain categorical statement, such as ‘very often’ to ‘not at all’. The preceding four weeks are used as the recall period, except for general health, which pertains to present status. The raw score for each CHQ-CF87 scale was based on the mean of valid items if at least half of the item in each scale was valid and transformed into a scale of zero (poor) to 100 (optimal), with higher scores indicating better functional health and well-being. A Norwegian version of this instrument with acceptable psychometric properties was used (17). Short Form Health Survey-36 (SF-36, first version) This instrument consists of 36 items across eight domains addressing physical and mental health. The domains are physical functioning (10 items describing ability to do physical tasks), role-physical (four items concerning difficulties or limitations in physical ability), bodily pain (two items), general health (five items), vitality (four items measuring level of energy), social functioning (two items about physical and mental health interfering with normal social life), role-emotional (three items about effects of mental problems on everyday life) and mental health (five items measuring moods). Except for the two role-functioning scales, which have dichotomised response choices, the responses are rated along three to six-point Likert-type scales with the preceding four weeks as the recall period, except for physical functioning and general health, which pertain to current status. The raw score for each SF-36 scale was based on the mean of valid items if at least half of the item in each scale was valid, and then transformed into a
©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2015 104, pp. 948–955
949
B atsvik et al.
Health-related quality of life after extremely preterm birth
Table 1 Demographic and clinical characteristics at 17 and 24-years-of-age in subjects born extremely preterm (EPB) with or without severe disability, and at term (TB)
Birth characteristics Gender, male/female, n Gestational age, weeks, mean (SD) Birth weight, gram, mean (SD) Age first follow-up, years, mean (SD) Age second follow-up, years, mean (SD) Neurosensory impairments, n Disabling CP Wheelchair Non-disabling CP Deaf/hard of hearing Blind/visually impaired Parents’ education Mother’s highest education, n (%) Primary school Upper secondary College/University Father’s highest education, n (%) Primary school Upper secondary College/University School and school support at 17 years, n (%)‡ Mainstream school School psychological counseling Academic assistance Education and employment at 24 years, n (%) Highest educational attainment Primary/secondary school College/University, four years Employment Still in education Working Redundancy Disability pension Living arrangements at 17 years/24 years, n Parents’ home§ Sheltered housing/institution Independent living With spouse or partner
EPB severe disability (n = 9)*
EPB healthy (n = 37)**
TB controls (n = 46)***
6/3 27.0 (1.9) 946 (176) 17.9 (1.7) 24.4 (1.8)
19/18 27.4 1029 17.6 24.2
25/21 (1.3) (194) (1.1) (1.2)
EPB severe disability vs TB controls p-value†
EPB healthy vs TB controls p-value†
3441 (310) 17.8 (1.2) 24.6 (1.2)
4 3 3 2 7
0 0 4 1 2
0 0 0 0 0
3 (43) 2 (29) 2 (29)
7 (24) 12 (41) 10 (35)
4 (11) 16 (43) 17 (46)
0.319
0.302
3 (43) 3 (43) 1 (14)
9 (30) 11 (37) 10 (33)
5 (14) 14 (39) 17 (47)
0.419
0.272
7 (78) 7 (78) 6 (67)
35 (95) 23 (62) 18 (49)
42 (91) 16 (35) 8 (17)
0.333 0.017 0.004
0.251 0.012 0.002
7 (88) 1 (12) 0
17 (49) 14 (40) 4 (11)
14 (35) 14 (35) 12 (30)
0.021
0.032
2 (25) 1 (12)
7 (20) 24 (69)
0.003
0.042
5 (63)
4 (11)
14 (35) 25 (63) 1 (3) 0
0.006/0.104
0.808/0.111
6/0 3/1 0/6 0/1
34/6 0/1 2/13 1/15
43/4 0/0 3/17 0/19
All information about demographic and clinical characteristics were obtained from a general questionnaire. *8 at second follow-up, **35 at second follow-up, ***40 at second follow-up. †Pearson’s chi square test. ‡Questions about learning difficulties were reported by parents. § With one parent or both. The bold values are statistically significant.
scale from 0 to 100, with higher scores indicating better functional health and well-being. A validated Norwegian version with acceptable reliability and validity was used (18). Health Behaviour in School-aged Children–Symptom Check List The HBSC-SCL questionnaire that was used at both measurement points assesses frequency of subjective health complaints during the previous six months. It is an eight-item scale addressing physical complaints, such as
950
headache, abdominal pain, back pain and dizziness (somatic subscore), and psychological complaints, such as feeling low, nervous, irritable and difficulty falling asleep (psychological subscore). Ratings were made on a five-point frequency scale with ‘rarely or never’ as zero and ‘approximately every day’ as four. An overall sum score of 0 to 32 and two subscores of 0–16 were generated, and lower scores indicated better subjective physical or psychological health. In this study, the subscores were used. Adequate reliability and validity have been reported in a Norwegian population of adolescents (19).
©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2015 104, pp. 948–955
B atsvik et al.
Health-related quality of life after extremely preterm birth
Ethics The Regional Committee on Medical Research Ethics of the Western Norway Health Region Authority approved the study. Informed written consent was obtained from each participant and from parents at the first follow-up if the subjects were minors. Statistical analyses For comparison of demographic data on a categorical level, differences between groups were analysed using exact chisquared tests. For the CHQ-87, SF-36 and the HBSC-SCL scales, one-way analyses of variance (ANOVA) using Tamhane corrections for multiple testing in post hoc analyses were applied for comparison between the three groups. Results are presented as group means (SD) and mean group differences with 95% confidence intervals (CI). Linear mixed effects models were used to compare development on the HBSC-SCL scores from 17 to 24 years of age between the three groups, including a time by group interaction in unadjusted and adjusted analyses for gender and education at the age of 24 years. p ≤ 0.05 was considered statistically significant. SPSS version 21 (SPSS Inc, Chicago, IL, USA) was used for ANOVA while the R (The R Foundation for Statistical Computing, Vienna, Austria) package nlme was used for mixed effects models.
RESULTS At the first assessment, 46 of 51 eligible EPB subjects (90%) participated. An average of 1.4 TB adolescents was invited for each EPB subject to recruit the matched control group. At the second follow-up, 43 EPB young adults and 40 TB controls (84% and 87%, respectively) participated (Fig. 1). All were Caucasians and native Norwegians.
The first follow-up at 17 years of age Group characteristics Similar proportions of healthy EPB and TB subjects (94%) and 67% of EPB subjects with severe disabilities were living at home with their parents (Table 1). The remaining EPB subjects with disabilities were living in some kind of sheltered housing. Most subjects were attending mainstream secondary schools. Of the healthy EPB adolescents, 49% had received academic support compared to 17% of the TB controls (p = 0.002), and 62% vs 35% (p = 0.012) had received support from the School Psychological Service. Health-related quality of life (CHQ-CF87) From the ANOVA analyses, EPB adolescents with severe disabilities reported lower scores on most domains, but only physical functioning was significantly lower compared to their TB peers (p = 0.034) (Table 2). The healthy EPB subjects and TB controls did not differ significantly on the physical and psychological domains. Subjective health complaints There were no significant differences between the healthy EPB subjects, the EPB subjects with severe disabilities and the TB controls in self-rated somatic and psychological health complaints, as measured by the HBSC-SCL in the ANOVA analyses (Table 2). The second follow-up at 24 years of age Group characteristics Most of the EPB and TB young adults were living independently or together with spouses or partners (Table 1). The EPB subjects with severe disabilities were mainly living independently of their families, while receiving some kind of care services or domestic help.
Eligible infants born extremely preterm and admitted to NICU* care in 1982 – 85 (n = 81)
Control infants born at term
First follow-up at age 15-19 years: 46/51 (90%)
First follow-up at age 15- 19 years: (n = 46)
Died in NICU or after discharge (n = 30) Declines, n = 5
Declines, n = 3
Declines, n = 3
Second follow-up at age 21- 25years 43/51 (84%)
Second follow-up at age 21 to 25 40/46 (87%)
Figure 1 Recruitment of subjects. *Neonatal intensive care unit.
©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2015 104, pp. 948–955
951
B atsvik et al.
Health-related quality of life after extremely preterm birth
Table 2 Self-reported functional health and well-being at adolescence in subjects born extremely preterm (EPB) with or without severe disability and controls born at term (TB) provided as mean scores (standard deviation) on the Child Health Questionnaire-Child Form 87 (CHQ-CF87) and the Health Behaviour School-aged Children-Symptom Checklist (HBSC-SCL) EPB severe disability (n = 9) CHQ-CF87 domain Physical functioning Role - physical Bodily pain General health Self-esteem Behaviour Role - behavior Mental health Role - emotional HBSC-SCL variables Somatic complaints Psychological complaints
Mean (SD) 52.3 (39.4) 81.5 (32.9) 61.1 (35.5) 63.2 (18.8) 71.2 (14.3) 73.5 (11.0) 96.3 (11.1) 75.0 (16.2) 81.5 (33.8) Mean(SD)min-max 2.4 (2.7) 0–6 3.4 (3.7) 0–12
EPB healthy (n = 37)
TB controls (n = 46)
94.5 95.8 73.8 67.0 70.2 80.1 97.3 74.6 86.5
95.1 96.1 73.0 66.7 70.6 79.6 98.6 75.6 94.0
(9.5) (9.9) (27.2) (17.9) (15.5) (12.5) (7.1) (15.8) (23.7)
3.2 (3.6) 0–13 2.9 (3.3) 0–12
(7.7) (10.8) (25.5) (17.7) (12.3) (11.3) (6.9) (12.6) (13.8)
2.9 (3.1) 0–12 3.0 (2.9) 0–14
EPB severe disability vs TB controls Mean difference (95%CI) 42.8 ( 82.3; 3.3)* 14.7 ( 47.5;18.3) 11.9 ( 47.7;23.8) 3.5 ( 22.7;15.6) 0.6 ( 13.8;15.1) 6.1 ( 17.3;5.1) 2.3 ( 13.4;8.9) 0.6 ( 17.0;15.7) 12.5 ( 46.3;21.4) Mean difference (95%CI) 0.4 ( 3.2;2.4) 0.4 ( 3.3;4.2)
EPB healthy vs TB controls
0.59 0.3 0.7 0.2 0.4 0.5 1.3 0.9 7.5
( ( ( ( ( ( ( ( (
5.3;4.1) 5.9;5.2) 13.5;15.0) 9.3;9.8) 8.1;7.2) 6.0;7.0) 5.0;2.5) 8.8;6.8) 18.3;3.4)
0.4 ( 1.5;2.2) 0.1 ( 1.8;1.6)
Differences between groups were analyzed with ANOVA using Tamhane correction for multiple testing in post hoc analyses. The bold value is statistically significant,*p = 0.034. Mixed effects models were used to assess differences in the HBSC-SCL scores between the three groups.
The TB subjects had achieved significantly higher education, and a higher proportion were still in education compared to the healthy EPB subjects (p = 0.042). Five of the eight EPB subjects with severe disabilities and three of the 37 healthy EPB subjects were receiving a disability pension, but none of the TB controls were. Health-related quality of life (SF-36) In the ANOVA analyses, the healthy EPB subjects scored consistently lower on all SF-36 domains and significantly lower on social functioning, emotional role and mental health than the TB controls (Table 3). The EPB subjects with disabilities tended to score lower than the TB controls on most of the scales, but the differences were not significant. The EPB subjects, as a whole, scored significantly lower than the TB controls on five domains of the SF-36. These five domains were bodily pain (mean differences 10.8, 95% CI 12.3 to 0.2), vitality (mean difference 8.9, 95% CI 16.3 to 1.5), social functioning (mean difference 13.3, 95% CI 23.7 to 2.9), emotional role (mean difference 22.8, 95%CI 39.0 to 6.6) and mental health (mean difference 9.6, 95%CI 16.1 to 3.1). Developmental changes in subjective health complaints Due to missing values in the adjustment variables, the adjusted analyses were based on 81 people compared with 92 people in the unadjusted analyses. In the unadjusted mixed effects analysis regarding the psychological health complaint, there was a significant group by time interaction (p = 0.003). The group differences were nonsignificant at adolescence (p = 0.867), but significant at young adulthood (p = 0.005). Specifically, the healthy EPB group had a higher mean score than the TB controls (2.5, 95% CI 1.0 to 4.0, p = 0.001). There were no
952
significant differences between the EPB subjects with disabilities and the TB controls at either age (p ≥ 0.131). For the healthy EPB subjects, the psychological complaint increased from 17 to 24 years of age (1.9, 95% CI 0.8 to 3.0, p = 0.001) and descriptive information is given in Tables 2 and 3 and Fig. 2. In the adjusted analyses regarding psychological health complaint (data not reported in Tables), the interaction was still significant (p = 0.005). The group differences were insignificant at adolescence (p = 0.883), but significant at young adulthood (p = 0.004). Specifically, the healthy EPB subjects had a higher mean score at young adulthood than the TB controls (2.7, 95% CI 1.1–4.2, p = 0.001). There were no significant differences between the EPB subjects with disabilities and the TB controls when they were adolescents or young adults (p ≥ 0.152). For the healthy EPB subjects, the psychological complaint increased (1.8, 95% CI 0.7–3.0, p = 0.002). In the unadjusted mixed effects analysis regarding the somatic health complaint, there was no significant group by time interaction (p = 0.135). Furthermore, there were no significant group differences at any measurement points (p ≥ 0.275) and no significant time differences in any group (p ≥ 0.169). In the adjusted analysis, there was no significant group by time interaction (p = 0.182). Furthermore, there were no significant group differences at any assessment points (p ≥ 0.271) and no significant time differences in any group (p ≥ 0.191).
DISCUSSION At young adulthood, but not at adolescence, the EPB subjects had more psychological complaints and rated their psychosocial HRQoL as inferior to that of their TB peers. In contrast to the TB controls, psychological complaints
©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2015 104, pp. 948–955
B atsvik et al.
Health-related quality of life after extremely preterm birth
Table 3 Self-reported health-related quality of life at 24-years-of-age in subjects born extremely preterm (EPB) with or without severe disabilities and controls born at term (TB) given as mean (standard deviation) scores on the Short Form Health Survey-36 (SF36) and the Health Behaviour School-aged Children-Symptom Checklist (HBSC-SCL) EPB severe disability (n = 8) SF-36 domain Physical functioning Role – physical Bodily pain General health Vitality Social functioning Role – emotional Mental health HBSC-SCL variables Somatic complaints Psychological complaints
Mean (SD) 56.9 (43.0) 68.8 (37.2) 73.9 (28.1) 65.9 (22.2) 53.1 (14.4) 87.5 (17.7) 83.3 (35.6) 75.5 (12.7) Mean (SD)min-max 3.8 (3.0) 0–9 4.0 (2.1) 1–7
EPB healthy (n = 35)
TB controls (n = 43)
92.1 80.5 68.4 74.1 50.7 74.3 61.4 69.5
94.2 86.9 80.2 77.1 60.0 90.0 88.3 80.2
(15.0) (30.4) (25.0) (25.0) (20.5) (30.2) (42.2) (17.8)
3.8 (3.9) 0–16 4.9 (4.5) 0–16§
(11.3) (25.9) (23.0) (20.0) (13.8) (17.3) (31.6) (12.2)
2.4 (2.2) 0–9 2.2 (2.3) 0–12
EPB severe disability vs TB controls Mean difference (95%CI) 37.4 ( 84.7; 9.9) 18.1 ( 59.1;22.8) 6.3 ( 27.2; 24.6) 11.3 ( 35.8; 13.3) 6.9 ( 22.8; 9.0) 2.5 ( 22.1; 17.1) 5.0 ( 44.3; 34.3) 4.7 ( 18.8; 9.4) Mean difference (95%CI) 1.4 ( 1.9; 4.7) 1.8 ( 0.5; 4.1)
EPB healthy vs TB controls
2.2 6.4 11.8 3.0 9.3 15.7 26.9 10.7
( ( ( ( ( ( ( (
9.7;5.4) 22.5; 9.7) 25.4;1.8) 15.9;9.9) 19.4;0.8) 30.0; 1.5)* 48.3; 5.5)† 19.5; 1.9)‡
1.4 ( 0.4;3.3) 2.5 (1.0;4.0) ¶
Differences between groups were analyzed with ANOVA using Tamhane correction for multiple testing in post hoc analyses. *p = 0.026, †p = 0.009, ‡p = 0.011.Mixed effects models were used to assess differences in the HBSC-SCL scores between 17 to 24-years-of-age, and between the three groups at each time, including a group by time interaction. EPB healthy between 17 to 24-years-of-age §p = 0.002. ¶p = 0.001. The bold values are statistically significant.
Haealth complaints (mean)
6
EP severe disability
EP healthy
Term born
5
*
4
**
3
2
1
0 Somatic complaints 17 yr
Somatic complaints 24 yr
Psychological complaints 17 yr
Psychological complaints 24 yr
Figure 2 Development of self-reported somatic and psychological health complaints from 17 to 24 years of age in the extremely preterm-born (EPB) subjects with or without severe disabilities and the term-born (TB) controls according to mean scores on the HBSC-SCL. Higher scores indicate more complaints. Mixed effects models, including group by time interaction, were used to assess differences from 17 to 24 years of age and differences between the three groups at each age of assessment. *p = 0.002 compared to 17 years of age, **p = 0.001 compared to TB peers at 24 years of age.
increased from adolescence to young adulthood. The negative trajectory was particularly noticeable for the healthy EPB subjects. The major strengths of this study were the high participation rate at both assessments and the recruitment of controls in a way that limited bias. A limitation was the relatively low number of participants, rendering the study at a particular risk of type-two errors, for failing to detect differences that may have been present. This limitation was particularly relevant when performing subgroup comparisons, such as
those involving participants with severe disabilities. A priori power calculation could not be performed as the distribution of the CHQ-CF87-scores in EPB subjects was unknown when planning the study. According to statistical advice, post hoc power calculations were not carried out. Instead, we quantified uncertainty by stating 95% confidence intervals and p-values. SF-36 is considered to be the adult counterpart to the CHQ-CF87 as it assesses similar domains of HRQoL (20), but direct comparison of the developmental trajectories could not be performed as the instruments are not constructed in the same way. Consistent with our results, earlier studies have demonstrated that EPB adolescents tend to rate their HRQoL as equal to that of TB peers despite extensively reported increased occurrence of academic underachievement, behavioural problems and various health problems (8,10). In the EPB adolescents with severe disabilities, physical limitation was not accompanied by significantly more limitations in social roles, or poorer psychological health, suggesting that physical ability was not considered an important aspect of HRQoL. These findings may indicate an optimistic attitude to life, ability to cope with life challenges, resilience or adjustment of expectations or changes in internal standards and values, even for those with severe disabilities. This is in line with some previous reports (21,22). There are, however, worrying reports of psychosocial and educational vulnerabilities among prematurely born subjects in late adolescence that may complicate their transition to adult life (8,23,24) and cause less resilience and increased risk of functional problems as adults (9,25). The main outcome of this study may support such concerns. In line with this, a previous longitudinal study demonstrated a nonsignificant tendency towards decrease in the HRQoL of prematurely born adults (13), and another study reported a small decrease for both EPB and TB young adults (22).
©2015 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2015 104, pp. 948–955
953
B atsvik et al.
Health-related quality of life after extremely preterm birth
Using SF-36, young adults with birthweights of
