BACKGROUND: This study investigated change of health-related quality of life (HRQL) in very preterm/very low birth weight (VP/VLBW; born at <32 weeks’ gestation and/or <1500 g birth weight) individuals from adolescence to adulthood. Are perceptions similar by different informants (self, parents) and is HRQL related to economic and social functioning?
METHODS: In a prospective whole-population sample in South Germany, 260 VP/VLBW and 229 term born individuals were assessed from birth to adulthood. HRQL was evaluated by self and parent report at age 13 and 26 years with the Health Utilities Index Mark 3 (HUI3), and economic and social functioning from interview and standard assessments at 26 years.
RESULTS: At both time points, HUI3 scores of VP/VLBW were reported to be lower compared with term born controls by participants and parents. Except for adolescent self-reports (P = .13) these differences were all significant (P < .05). In contrast to participants themselves, parents reported VP/VLBW individuals’ HRQL to be worsening over time (change of mean HUI3 scores: 0.88–0.86, P = .03). Parents, particularly, reported negative changes in emotion and pain for VP/VLBW individuals over time. Participant and parent-perceived HRQL was negatively related to economic and social functioning outcomes such as receiving social benefits, unemployment, dating romantic partner or having friends.
CONCLUSIONS: VP/VLBW individuals and their parents perceive HRQL to be lower compared with term controls in adolescence and in adulthood. Lower HRQL was related to economic and social functioning problems in adulthood. No evidence for improvement of HRQL into adulthood was found in this geographical sample in Germany.
- ELBW —
- extremely low birth weight
- HRQL —
- health-related quality of life
- HUI3 —
- Health Utilities Index Mark 3
- MAU —
- multiattribute utility
- SES —
- socioeconomic status
- VP/VLBW —
- very preterm/very low birth weight
What’s Known on This Subject:
Very preterm individuals are at risk for health problems and lower health-related quality of life (HRQL) in childhood. However, by adulthood those born very preterm may perceive their HRQL similar to their term born counterparts.
What This Study Adds:
Those born very preterm perceive their HRQL as lower than their parents and as poorer than term born controls in adolescence and adulthood. Lower HRQL is related to lower economic functioning and poorer social relationships with peers and partners.
Very preterm (VP; <32 weeks’ gestation) or very low birth weight (VLBW; <1500 g) born infants are at increased risk of long-term physical and psychological sequelae including functional limitations and chronic health disorders.1 In contrast, health-related quality of life (HRQL) refers to the impact of health on an individual’s overall psychological, social, and physical well-being2,3 and HRQL assessments are widely used to determine the impact of health care.2 They provide comprehensive and multidimensional information on subjective health and functional status perceived by the preterm born individuals themselves2 and thus convey valuable information for researchers and health professionals in addition to objective assessments of health and function.4
Past research suggests that preschool and school-aged VP/VLBW infants have poorer HRQL than term born comparisons.5,6 However, despite evidence of a continuing increased risk of disabilities, preterm born adolescents and young adults reported not significantly lower HRQL compared with term born counterparts.2,7–9 As of yet, only 2 longitudinal studies reported on change of self-perceived HRQL from adolescence to adulthood.2,3,10,11 The findings are contradictory, 1 revealed no change in HRQL over time10 and the other a decrease.3 Furthermore, how HRQL relates to economic functioning such as employment, independent living, and social relationships in adulthood in preterm born individuals has not been previously examined.
The current study investigated (1) whether overall HRQL improves or worsens over time (from adolescence to adulthood) for VP/VLBW individuals; (2) whether changes over time are similar in participant and parent reports; (3) which attributes are affected; and (4) whether poorer HRQL is related to adult economic and social relationship functioning (ie, the ability to lead an independent and self-sufficient life in adulthood).
The Bavarian Longitudinal Study is a geographically defined prospective whole population sample of children born in Southern Bavaria (Germany) between January 1985 and March 1986 who required admission to 1 of the local 16 children’s hospitals within the first 10 days after birth (N = 7505; 10.6% of all live births).12,13 Of this cohort, 682 were born VP/VLBW, 411 of these were eligible for the 26-year follow-up assessment, and 260 (63.3%) participated. Healthy term infants born in the same obstetric hospitals were recruited as controls. Of the initial 916 control children alive at 6 years, 350 were randomly selected as term controls within the stratification variables gender and family socioeconomic status (SES) to be comparable to the VP/VLBW sample. Of these, 308 were eligible for the 26-year follow-up assessment, and 229 (74.4%) participated (Fig 1).
Across both assessment waves at 13 and 26 years, data on health utility function reported by the participants themselves were available for 214 VP/VLBW and 201 term controls.
Parent reports on their offspring’s health utility function were available for 197 VP/VLBW and 204 term born controls for both time points. Additionally, parents as proxy informants provided information for 13 impaired VP/VLBW individuals who were not able to be interviewed themselves or to fill in questionnaires.
In total, 190 VP/VLBW and 201 term control participants had full data sets across both assessments (age 13 and 26 years). Ethical approval was obtained from the University of Munich Children’s Hospital and the Bavarian Health Council (Landesärztekammer Bayern). Ethical approval for follow-up in adulthood was granted by the Ethical Board of the University Hospital Bonn (reference # 159/09). Informed written consent was provided initially by parents within 48 hours of their child’s birth and all participants gave fully informed written consent for the assessments in adulthood. In case of severe impairment of the adult participant, consent was provided by an assigned guardian (usually the parents).
Health Status and HRQL at 13 and 26 Years of Age
Participants and their parents separately completed the unedited 15-item Health Utilities Index (HUI) questionnaire for health status assessment.14 Responses were mapped onto the Health Utilities Index Mark 3 (HUI3) health status classification system. The HUI3 covers 8 attributes: vision, hearing, speech, ambulation, dexterity, emotion, cognition, and pain. Function within each of these attributes is graded from normal function (level 1) to severe impairment (levels 4–6). Health states were converted into multiplicative multiattribute utility (MAU) scores using a published utility algorithm on the basis of preferences of a randomly selected general population sample of Canadian adults.14 Thus, MAU scores indicate HRQL of adolescents and adults on the basis of societal standards (in this case, Canadian adults). MAU scores can range from −0.36 to 1.00 (representing a health state with the lowest level and the highest level of function across all attributes) on an interval scale.
Economic and Social Functioning Outcomes
Adapted from several established and widely used instruments, such as the German Socioeconomic Panel Study,15,16 and the Avon Longitudinal Study of Parents and Children,17 participants were interviewed about their economic and social functioning (ie, receiving social benefits, unemployment, housing situation, emotional support of parents, dating romantic partner, friends, and social activities).
Gestational age, birth weight, and gender were obtained from perinatal records at birth.12 Family SES at birth defined as low, middle, or high.18 Disability was determined in childhood: suffering from grade 3 or 4 cerebral palsy,19 blindness, or deafness.20 Cognitive impairment was defined as a IQ < 2 SD assessed with the German version of the Kaufman Assessment Battery for Children at 8 years.21,22
All analyses were conducted by using SPSS Version 22 (IBM SPSS Statistics, IBM Corporation). Differences between VP/VLBW and term control adults were tested with analysis of variance or χ2 test.
Changes in health utility function (HUI3 MAU score) from adolescence to adulthood were tested with paired t tests. Changes in HRQL scores from adolescence to adulthood were judged as clinically important when they reached a difference of 0.03.23 To test for interaction effects, repeated measure analyses of covariance (in between factor grouping [VP/VLBW versus term-born] and covariates gender and SES) were applied. Effect sizes (partial η squared [ηp2]) were interpreted according to Cohen’s guidelines: 0.02 = small, 0.13 = medium, 0.26 = large. Stability of HUI3 MAU scores were assessed with Pearson correlations and groups differences tested with Fisher’s z test.
HUI3 MAU scores were categorized into 4 disability levels (no disability = 1.00; mild disability = 0.89–0.99; moderate disability = 0.70–0.88; severe disability = less than 0.70)24 to assess changes over time.
Function within each of the 8 HUI3 attributes was recorded as suboptimal if a level of functional impairment (level 2 or above)4,25 was reported. Changes over time of suboptimal function of attributes were tested with McNemar tests.
Additionally, differences in the number of single attributes that were suboptimal between groups (VP/VLBW versus term-born) and their change over time were tested by using Mann–Whitney and Wilcoxon tests, respectively.
To determine whether health utility scores were related to economic and social functioning, adult HUI3 MAU scores were regressed on measures of economic (ie, receiving social benefits, unemployment, and not living independently) and social functioning (ie, emotional support from parents, dating romantic partner, and having friends).
By definition, VP/VLBW participants were born with lower gestational age and birth weight than term born controls but no group differences were found with regard to gender (Table 1). However, more VP/VLBW adult participants had been born into families of lower and middle SES than term born individuals. Family SES and gender25 were controlled for in subsequent multivariate analyses.
HRQL in VP/VLBW and Term Born Individuals: Changes Over Time
At both time points, HRQL of VP/VLBW individuals were reported by participants and their parents to be poorer compared with term born controls, and when parental proxy reports for severely impaired VP/VLBW adults were included the difference between HUI3 MAU scores of VP/VLBW and term controls further increased (Fig 2). However, based on adolescents’ self-reported HUI3 MAU scores, VP/VLBW individuals did not have significantly lower scores compared with term born controls (P = .13; all other Ps < .05).
Self-reports of VP/VLBW individuals’ HRQL did not change from adolescence to adulthood (mean HUI3 MAU score at both time points 0.86, P = .99). Similarly, term born controls HRQL from adolescence (mean HUI3 MAU score 0.88) to adulthood (mean HUI3 MAU score 0.89) did not change significantly (P = .34). According to parents’ reports, VP/VLBW individuals’ HRQL decreased over time (mean HUI3 MAU score: 0.88 in adolescence and 0.86 in adulthood, P = .03). When parent proxy-reports on VP/VLBW individuals were included, mean HUI3 MAU scores were even lower (0.85 in adolescence and 0.82 in adulthood, P = .02). In contrast, parent-reported HRQL for term control participants increased significantly from adolescence to adulthood (mean HUI3 MAU score 0.92–0.94, P = .04.
The repeated analysis of variance confirmed a significant interaction effect between time and group in HRQL (P = .003). The change over time was significantly different on parent-reported HRQL depending on group membership (VP/VLBW and term control). However, the effect size was small (ηp2 0.023). The results were similar when parent proxy-reports on VP/VLBW individuals were included.
Pearson correlation scores over time on the HUI3 MAU score revealed positive and significant correlations between 13- and 26-year self (VP/VLBW r = 0.475, term controls r = 0.203) and parent reported (VP/VLBW including proxy-cases r = 0.825, VP/VLBW excluding proxy-cases r = 0.675, term control r = 0.152) HUI3 MAU scores. Correlations and thus stability of HUI3 MAU scores were higher for VP/VLBW than term controls in self (z score 3.00, P = .003) and parent reports (z scores 6.50 and 10.09, respectively excluding and including proxy-cases, Ps < .001).
Considering HUI3 MAU scores in 4 levels of disability (none, mild, moderate, severe), more VP/VLBW individuals worsened from adolescence to adulthood and were less likely to improve compared with term born adults. Reports by participants and parents yielded similar results, but group differences were larger in parent than participant reports (Table 2).
Group differences and changes over time in the number of attributes that were suboptimal are presented in Table 3. Group differences in parent reports were found at both time points. Parents also reported more suboptimal attributes for their VP/VLBW offspring in adulthood.
Which Attributes of HRQL Are Affected?
There were significant changes in suboptimal function of vision (ie, “able to see well enough to read ordinary newsprint and recognize a friend on the other side of the street, but with glasses” to “unable to see well enough to read ordinary newsprint and unable to recognize a friend on the other side of the street, even with glasses”) and speech in both VP/VLBW and term control individuals. Suboptimal function in vision increased, whereas suboptimal function of speech decreased over time. Suboptimal function of cognition (ie, “able to remember most things, but has a little difficulty when trying to think and solve day to day problems” to “very forgetful, and has great difficulty when trying to think or solve day to day problems”) decreased for term control adults but not for VP/VLBW adults.
Similar to participants’ self-reports, suboptimal function of vision worsened in VP/VLBW and term born individuals. Suboptimal functioning of speech decreased in both groups but was significant only for term control adults. In contrast, VP/VLBW adults’ suboptimal function of emotion increased over time and decreased for term control adults. As reported by participants, suboptimal function of cognition decreased over time but was significant only for term born adults. Similar to emotion, suboptimal pain increased in VP/VLBW individuals and decreased in term born adults.
Economic and Social Functioning and HRQL in Adulthood
Economic and social functioning measures were correlated with HRQL in adulthood (Table 5). Of the economic functioning measures, receiving social benefits and unemployment was negatively associated with participant and parent-reported HRQL in adulthood. Not living independently (ie, living at parents’ house or in a home) was associated with parent-reported HRQL but not with participant-reported HRQL. Regarding social functioning measures, having dated a romantic partner and having many friends was related to both, participant and parent-reported HRQL in adulthood. In addition, emotional support by parents (ie, exchange of thoughts and feelings with parents) was significantly associated with participant-reported HRQL and social activities with parent-reported HRQL. Except for the correlation of unemployment with self-reported HRQL and social activities with parent-reported HRQL, all associations remained significant after adjusting for each other, prematurity, gender, and SES at birth (Table 5).
Frequencies of economic and social functioning measures and differences between groups are shown in Supplemental Table 7.
This prospective whole-population study revealed that compared with term controls, VP/VLBW individuals had lower HRQL assessed by HUI3 MAU scores in adulthood as reported by participants and parents. Parents reported HRQL of their VP/VLBW offspring to be worsening from adolescence to adulthood, whereas participants themselves reported no change. Additionally, stability of HUI3 MAU scores was significantly higher for VP/VLBW than term controls in parent and self-reports. Notably, parents reported worsening in emotion and pain for VP/VLBW individuals. Both participant- and parent-reported HRQL were related to economic and social functioning outcomes.
Preterm infants are at increased risk of health problems and functional limitations1 that often continue into adulthood. Despite this, premature born adults have previously reported similar HRQL compared with term born peers.2,7–9 To our knowledge, this is the first prospective study of VP/VLBW adults, reporting on change of HRQL perceived by participants themselves and their parents.
Comparable with most studies examining HRQL in preterm born adolescents,7,26 we found no differences between VP/VLBW and term individuals in self-reports in adolescence. However, our findings partly differ from the other 2 previous prospective studies that asked preterm participants about their HRQL in adolescence and again in adulthood. A Dutch study revealed a nonsignificant change of HRQL from age 14 to 19 years,11 as well as from age 19 to 28 years10 for VP/VLBW individuals consistent with self-reports by VP/VLBW and term controls in this German sample. Although we detected significant differences in self-reports of HRQL in adulthood, these were not found in the Dutch sample. The study in the Netherlands had no term control group but used published norms for comparison and had highly selective attrition in their online survey that may make direct comparisons difficult.
The Canadian study3 followed a cohort of extremely low birth weight (ELBW; birth weight < 1000 g) and normal birth weight individuals and revealed a significant decrease of health utility scores in both ELBW and normal birth weight adults but no group differences. Decreases from adolescence to adulthood in HRQL have been revealed in previous general population studies,27–29 and have been interpreted to reflect the challenges posed in the transition to adulthood such as career decisions, studies, living independently, or finding a partner. Direct comparison with the Canadian study may be impeded as it used directly measured preferences and a different health utility system (HUI2) than in this study. Furthermore, it included a more immature born population of ELBW individuals and cultural differences should also be considered.2,12
The unique contribution of our study is that in addition to participants themselves, parents were asked to report on their offspring’s HRQL. Furthermore, we were able to compare stability of HRQL for VP/VLBW individuals and term controls. We found that VP/VLBW HRQL was more stable in the transition to adulthood than for controls. Similar higher stability over time has been reported for objective functioning such as IQ30 indicating that life trajectories may be more fixed in preterm than term born populations.
Parents of VP/VLBW individuals reported significantly lower HUI3 MAU scores than parents of term controls, both, in adolescence and adulthood, in particular, when proxy-cases were included. Different from participants themselves, parents reported a decrease of mean HRQL score for VP/VLBW individuals but increasing scores for term control adults. However, the changes were small and below the 0.03 difference proposed to be of clinical relevance.23
Differences between self- and parent-reported HRQL have been observed and reported previously.7 Parents appear to have a different perspective on their offspring’s actual function and compare them with all peers,25 whereas participants’ perspectives on their HRQL may be based on factors that are related to socializing with their friends and peers. Parents also may be more concerned about their offspring’s health and general well-being31 and transition to independence than preterm born adults themselves. Moreover, previous literature has revealed that agreement between individuals and their parents is dependent on child age, parents’ own quality of life, and the attribute examined.32,33 High agreement was found in physical and observable attributes (eg, vision, hearing, ambulation), whereas for attributes that reflect internal processes and are not easily observable (eg, emotion, pain, cognition) agreement was low. Accordingly and similar to previous findings,3,10,11,34 the attribute with the most prominent decrease in function was vision, for both VP/VLBW and term control adults and reported by parents and participants alike. Further affected attributes, but reported by parents only, were emotion and pain; both attributes worsened for VP/VLBW individuals but improved for term control individuals.
Despite group differences in mean HRQL scores, we found similar and significant associations between both participant- and parent-reported HRQL and measures of economic and social functioning in adulthood in both VP/VLBW and term controls. Particularly, receiving social benefits, dating a romantic partner, and having many friends were significantly correlated with HRQL scores. Those born VP/VLBW had more often economic and social functioning problems (Supplementary Table 7).9 This may partly explain lower HRQL ratings of VP/VLBW individuals and their parents compared with term controls in adulthood. Thus, both self- and parent-reported measures of health status are related to real life outcomes, regardless of prematurity, gender, and SES. This may be important for planning and implementation of health and social services.
This study has a number of strengths. It is a long-term follow-up study of a large whole population sample of VP/VLBW and term-born individuals recruited in the same hospitals by using a valid, reliable, and widely accepted measurement tool of HRQL. Second, this study included participant and parent reports. In addition, we used proxy reports by parents for VP/VLBW adults who were too impaired to report on their own HRQL.
There are also limitations. First, although 63% and 74% of eligible VP/VLBW and term-born individuals, respectively, assessed in childhood could be assessed at 26 years, the dropout was not random. VP/VLBW and term-born individuals who were socially disadvantaged at birth were less likely to continue participation,35 a problem in many longitudinal studies36 and it may affect group comparisons.37 Nevertheless, simulations have shown that predictions only marginally change even when dropout is selective or correlated with the outcome.37 Second, as the HUI3 MAU score is an indirect measure of HRQL population preferences, a replication using measures of HRQL reflecting individual preferences may be considered.26 Third, the items used for the economic and social functioning measures were adapted from large socioeconomic panel and longitudinal studies. Although not fine-graded, they provide a view of severe economic or social outcomes affecting adults’ life chances.38
Those born VP/VLBW show little change in HRQL into adulthood, whereas their parents perceive a deterioration at this time of transition into adult roles. HRQL is strongly related to real life economic and social functioning outcomes. Our findings support assessments of subjective health and functional outcomes in addition to objective assessments of health and functioning in at-risk populations.
We thank all current and former Bavarian Longitudinal Study group members, pediatricians, psychologists and research nurses. Moreover, we thank those who contributed to study organization, recruitment, data collection, and management at the 26-year assessment: Barbara Busch, Stephan Czeschka, Claudia Grünzinger, Christian Koch, Diana Kurze, Sonja Perk, Andrea Schreier, and Julia Trummer. Special thanks are due to the study individuals and their families.
- Accepted January 12, 2016.
- Address correspondence to Dieter Wolke, PhD, Department of Psychology, University of Warwick, Coventry CV4 7AL, United Kingdom. E-mail:
The contents are solely the responsibility of the authors and do not necessarily represent the official view of the BMBF. Information on BMBF is available on http://www.bmbf.de/en/.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: This study was supported by grants PKE24, JUG14, 01EP9504, and 01ER0801 from the German Federal Ministry of Education and Science (BMBF).
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- Saigal S,
- Stoskopf B,
- Pinelli J, et al
- Verrips E,
- Vogels T,
- Saigal S, et al
- Riegel K,
- Ohrt B,
- Wolke D
- Furlong W,
- Feeny D,
- Torrance G, et al
- Frick JR,
- Jenkings SP,
- Lillard DR,
- Lipps O,
- Wooden M
- Wagner GG,
- Frick JR,
- Schupp J
- Bauer A
- Kaufman AS,
- Kaufman N
- Melchers P,
- Preuss U
- Breeman LD,
- Jaekel J,
- Baumann N,
- Bartmann P,
- Wolke D
- Saigal S,
- Stoskopf B,
- Boyle M, et al
- Hille ETM,
- Elbertse L,
- Gravenhorst JB,
- Brand R,
- Verloove-Vanhorick SP; Dutch POPS-19 Collaborative Study Group
- Wolke D,
- Waylen A,
- Samara M, et al
- Copyright © 2016 by the American Academy of Pediatrics