CONTEXT

There is a lack of research on individual perceptions of social experiences and social relationships among very preterm (VP) adults compared with term-born peers.

OBJECTIVE

To investigate self-perceived social functioning in adults born VP (<32 weeks’ gestation) and/or with very low birth weight (VLBW) (<1500g) compared with term-born adults (≥37 weeks’ gestation) using an individual participant data (IPD) meta-analysis.

DATA SOURCES

Two international consortia: Research on European Children and Adults born Preterm and Adults Born Preterm International Collaboration.

STUDY SELECTION

Cohorts with outcomes assessed by using the Adult Self-Report Adaptive Functioning scales (friends, spouse/partner, family, job, and education) in both groups.

DATA EXTRACTION

IPD from 5 eligible cohorts were collected. Raw-sum scores for each scale were standardized as z scores by using mean and SD of controls for each cohort. Pooled effect size was measured by difference (Δ) in means between groups.

RESULTS

One-stage analyses (1285 participants) revealed significantly lower scores for relationships with friends in VP/VLBW adults compared with controls (Δ −0.37, 95% confidence interval [CI]: −0.61 to −0.13). Differences were similar after adjusting for sex, age, and socioeconomic status (Δ −0.39, 95% CI: −0.63 to −0.15) and after excluding participants with neurosensory impairment (Δ −0.34, 95% CI: −0.61 to −0.07). No significant differences were found in other domains.

LIMITATIONS

Generalizability of research findings to VP survivors born in recent decades.

CONCLUSIONS

VP/VLBW adults scored their relationship with friends lower but perceived their family and partner relationships, as well as work and educational experiences, as comparable to those of controls.

Very preterm (VP) birth (<32 weeks' gestation) is associated with increased risk of long-term respiratory, neurologic, cognitive, and psychosocial problems.14  Furthermore, adverse impacts on social functioning in adulthood have been reported.5  Social functioning reflects an individual's interactions with human environments and the ability to fulfill their role within environments, such as education, work, and social relationships with friends, partners, and family.6  In 2 recent systematic reviews,5,7  researchers indicate that VP adults have lower educational attainment, higher rates of unemployment, and decreased likelihood of having a partner compared with term-born adults.

In most studies of VP adults, researchers assessed social functioning on the basis of factual information, such as educational attainment, income level, being employed (yes or no), receiving social benefits (yes or no), or being in a romantic relationship (yes or no).5,7  In contrast, there is a lack of research on VP adults’ individual perceptions, that is, how they perceive their experiences and relationships with others. Social functioning in preterm adults compared with term-born peers may differ depending on whether it is measured by using factual information or individual perceptions. For instance, although VP adults may be less likely to have romantic relationships, in previous studies, researchers suggest that once they are in a relationship, the quality of the relationship (eg, satisfaction, intimacy, or closeness) is perceived as similar compared with term-born peers.5  Little is known about VP adults’ perceptions of work and educational experiences (eg, satisfaction, stress, and getting along with colleagues and/or fellow students) when they are in employment or education.

Adults born VP or with very low birth weight (VLBW) (birth weight <1500g) report similar relations with family members compared to control participants813  but reported having fewer friends.8,9,14  There is less exploration of other aspects of friendship, such as getting along with friends and contact and/or visiting frequency.

Demographic, perinatal, and childhood factors may be associated with social functioning within VP and/or VLBW adults. Using factual information, compared with VP and/or VLBW male adults, more VP and/or VLBW female adults are reported to have reached higher education, but fewer are in employment.15,16  No differences exist in social outcomes (eg, cohabiting with a romantic partner) between singleton and multiple-birth VLBW adults.17  It is unclear whether sex and multiple birth relate to self-perceived social functioning. Associations are also unclear for socioeconomic status (SES) or neonatal complications, such as bronchopulmonary dysplasia (BPD) or intraventricular hemorrhage (IVH), themselves associated with adverse cognitive or motor outcomes.18  Furthermore, neurosensory disabilities diagnosed in childhood are associated with fewer friends and poorer peer relationships,14  whereas being born small for gestational age (SGA) is not associated with poorer family and/or friend relations.8  It is unknown whether these factors relate to other aspects of self-perceived social functioning.

Meta-analysis of individual participant data (IPD) is the gold standard to identify, appraise, and synthesize the evidence on the same research question from multiple studies by obtaining individual-level data on each participant from each eligible study.19,20  IPD meta-analysis has several advantages over traditional meta-analysis of aggregate data.21  It permits the use of consistent inclusion and exclusion criteria across studies and the incorporation of unpublished studies to avoid publication bias.19  Other advantages include using unified criteria to define participants and to harmonize data, standardizing the analysis across studies, ensuring sufficient power to examine the role of potential exposure variables, and the application of sophisticated modeling techniques.22 

We undertook an IPD meta-analysis combining individual-level data from prospective cohorts of VP or VLBW adults and term-born peers within 2 international consortia: Research on European Children and Adults born Preterm (RECAP Preterm) (https://recap-preterm.eu/) and Adults born Preterm International Collaboration (APIC) (https://www.apic-preterm.org/). Our objectives were, first, to investigate differences in self-perceived social functioning in VP and/or VLBW adults compared with term-born peers and, secondly, to examine whether childhood neurosensory impairment (NSI) and demographic and perinatal factors were associated with self-perceived social functioning among VP and/or VLBW adults.

This IPD meta-analysis followed the Preferred Reporting Items for a Systematic Review and Meta-Analysis of IPD reporting guideline23  and was registered with PROSPERO International prospective register of systematic reviews (CRD42020195610).

Cohorts were identified from the APIC and RECAP Preterm consortia, for which relevant data on self-perceived social functioning in adulthood were available. APIC and RECAP Preterm comprise cohorts from Europe, Australia, New Zealand, United States, and Canada. APIC is a research network aimed at studying health and well-being of adults born preterm through individual participant and aggregate data meta-analyses across multiple cohorts. In RECAP Preterm, data from European cohorts and other cohorts of APIC members are brought together, making it possible to increase the power and generalizability of research because data reflect broad geographic, cultural, and health system diversity. Seven adult cohorts within RECAP Preterm and 4 of the 6 contacted APIC members who agreed to collaborate were asked to provide data on adult outcome measures, childhood NSI, and demographic and perinatal factors. Inclusion criteria included (1) individuals born VP (<32 weeks' gestation) or VLBW (<1500 g), (2) a control group, (3) participants assessed in adulthood (mean age ≥18 years), (4) and self-perceived social functioning measured using the Achenbach System of Empirically Based Assessment Adult Self-Report (ASR) Adaptive Functioning scales.

To search for possible additional cohorts around the world, we additionally performed a systematic search for articles published in PubMed, PsycINFO, Web of Science, Scopus, and Embase on August 4, 2020. The following keywords were used: (preterm* OR “low birth weight” OR “low birthweight”) AND (adult*) AND (ASEBA OR ASR OR YSR OR “Achenbach System of Empirically Based Assessment” OR “Adult Self-Report” OR “Youth Self-Report” OR Achenbach* OR Adaptive Function*). Each potentially eligible study was assessed by 2 researchers independently, with disagreements resolved by discussion.

Anonymized data sets from eligible cohorts were transferred to the University of Warwick under signed grant agreements or data transfer agreements. Data included cohort-level information (eg, total number of participants and drop-out rates) and individual-level information: demographic and perinatal characteristics, childhood NSI, and self-perceived social functioning. Data sharing complied with the ethics approvals of local ethics committees of the separate studies. All assessed participants gave their informed consent.

Data were checked for internal consistency, and missing items and were cross-checked with published reports. Investigators were contacted if there was inconsistent or incomplete information. Item-level data were requested if scale-level data were not available. Each cohort was assessed on the basis of sample size, year of recruitment, assessment of outcome, and follow-up rates. Risk of bias of each cohort was assessed by 2 researchers independently by using the Newcastle-Ottawa Scale,24  with disagreements resolved by discussion. Scores on the scale range from 0 to 9, with higher scores indicating higher quality.

We used unified criteria to form preterm or control groups: participants born VP (<32 weeks' gestation) and/or VLBW (<1500g) were included in the VP and/or VLBW group and term-born (≥37 weeks' gestation) participants in the control group.

Demographic and perinatal variables included age at assessment, sex, SES, birth weight z score, multiple birth, BPD, and IVH. Birth weight z scores were calculated by using the Fenton growth chart.25  BPD definitions varied, but researchers used either the criterion of oxygen dependency at 36 weeks’ postmenstrual age or for >28 days after birth.26  IVH was classified according to Papile et al,27  but some cohorts provided either grades 3 and 4 or grades 2 and 3 combined. Thus, IVH was harmonized into no IVH versus any IVH, and a subanalysis compared no IVH or IVH grades 1 to 2 versus IVH grades 3 to 4 in cohorts in which this was possible. SES was measured by using maternal education collected at birth or in childhood and was classified by using the International Standard Classification of Education (ISCED)28 : (1) low, equivalent to ISCED 0 to 2; (2) medium, ISCED 3 to 5; and (3) high, ISCED 6 to 8. Childhood NSI was defined as ≥1 of the following: visual impairment (blind in 1 or both eyes), hearing impairment (not corrected by hearing aids), nonambulatory cerebral palsy, or cognitive impairment (childhood IQ <70 or other information indicating cognitive impairment).

Self-perceived social functioning was measured by using the ASR Adaptive Functioning scales, which include 5 domains (family, friends, spouse or partner, job, and education) and a total of 34 items.29  For each scale, higher scores indicate higher functioning. The family scale contains 9 items about which respondents are asked to rate how well they get along with various family members on a Likert scale (0 = worse than average, 1 = variable or average, and 2 = better than average). The friends scale yields a total score on the basis of 4 items (number of close friends, frequency of contacts with close friends, getting along with close friends, and visits by friends and/or family), each scored 0 to 3, with a maximum of 12 points. The spouse or partner scale comprises 8 items (rated 0, 1, or 2) dealing with how well respondents get along with their partner, share responsibilities, and enjoy similar activities; how satisfied they are with their partner; and how well they like their partner’s friends and family. The job scale comprises 8 items (rated 0, 1, or 2) dealing with how well respondents get along with co-workers and bosses and do their work, how satisfying or stressful they find their work, and how often they worry about work or do things that may cause them to lose their job. The education scale comprises 5 items (rated 0, 1, or 2) dealing with how well respondents get along with fellow students, achieve in their studies, and finish their work; how satisfied respondents feel with their educational situation; and how often respondents do things that may cause them to fail. Respondents who were not living with a spouse or partner, did not have a job, or were not in education during the preceding 6 months omitted these sections. Achenbach and Rescorla reported alphas of 0.51 to 0.78 for the Adaptive Functioning scales, and these alphas are reasonable for scales that have relatively few items.30  Raw-sum scores were calculated for the 5 scales within each cohort. Multicountry norms31  were not available for each country of the cohorts in this study. Raw-sum scores were converted to z scores by using mean and SD of control participants for each scale within each cohort.

IPD were harmonized to directly comparable variables. IPD meta-analyses were performed in Stata 16.0 (Stata Corp, College Station, TX). Along with a 2-stage approach32  specified in the published proposal, a 1-stage approach was also used33  because there remain debates regarding which approach is optimal.34 

Using the 2-stage approach, mean differences and 95% confidence intervals (CIs) for outcomes between the VP and/or VLBW and control groups were derived by using linear regression in each study separately. Pooled effect estimates were then derived from different studies by using the random-effects inverse variance method with the restricted maximum likelihood (REML) estimator.35  Pooled effect size was measured by unstandardized mean difference (Δ) in z scores because as outcomes were reported on a meaningful scale and all studies used the same scale.36  Heterogeneity was quantified by I2 and τ2.37  Low heterogeneity was defined as an I2 value <40% and considerable heterogeneity defined as a value >75%.

Using the 1-stage approach, we analyzed the IPD from all studies simultaneously using linear mixed-effects models with the REML estimator, which assumed a random group effect (VP and/or VLBW adults versus control adults). This approach also accounted for the clustering effect of participants in studies.33,34  The above procedure was repeated to estimate effect sizes after adjusting for covariates (sex, age at assessment, and SES), which were added as fixed effects. Between-study heterogeneity between the 2-stage and 1-stage approaches were compared by using τ2.34  As there is evidence indicating that differences in some domains of social functioning (eg, friends) between VP and/or VLBW and control adults were only found for female participants,10  sex and group interactions were also tested.

All analyses were repeated by excluding participants with childhood NSI. Analyses were then restricted to VP and/or VLBW participants only to investigate the effects of NSI and demographic and perinatal variables by using linear mixed-effects models. Corrections for multiple comparisons were applied by using the Benjamini-Hochberg procedure.38 

Five of the 11 participating RECAP Preterm and APIC adult cohorts were eligible for inclusion: the EPICure study (the United Kingdom and Ireland),39  the Preterm Birth and Early Life Programming of Adult Health and Disease Study (ESTER, Finland),40,41  the Helsinki Study of Very Low Birth Weight Adults (HESVA, Finland),42,43  the NTNU Low Birth Weight in a Lifetime Perspective study (NTNU LBW Life, Norway),8,9  and the Arvo Ylppö Longitudinal Study (AYLS, Finland).44  A literature search of published studies yielded only 2 cohort studies (NTNU and the Cleveland study in the United States10 ). NTNU published data on the friends and family scales at ages 20 and 23 years8,9 ; in this analysis, we used data from the most recent follow-up at age 26 years. In Fig 1, we illustrate the screening process of the eligible cohort studies. IPD were sought and obtained for the 5 eligible cohorts. Data for the Cleveland study were not available at the time of final analyses.

FIGURE 1

The Preferred Reporting Items for a Systematic Review and Meta-Analysis of IPD flow diagram detailing the screen process of the eligible cohort studies for the pooled analysis. a Data for the Cleveland study were unfortunately not available at the time of final analyses.

FIGURE 1

The Preferred Reporting Items for a Systematic Review and Meta-Analysis of IPD flow diagram detailing the screen process of the eligible cohort studies for the pooled analysis. a Data for the Cleveland study were unfortunately not available at the time of final analyses.

Close modal

All cohorts were followed from birth to adulthood, with drop-out rates ranging from 29.2% to 62.1% (Table 1). They scored 7 to 8 on the Newcastle-Ottawa Scale (Supplemental Table 4). The EPICure study was national, comprising births in 1995, compared with other regional cohorts with births in the 1970s and 1980s. The VP and/or VLBW and control groups within each cohort were born during the same period. We used unified criteria for VP and/or VLBW and control groups, which may differ from the original inclusion criteria for the index children in each cohort (Table 1).

TABLE 1

Summary of the 5 Adult Cohorts Included in the IPD Meta-analysis

CohortCountryScaleBirth YearMean Age of Assessment, yOriginal Index Eligibility CriteriaInitial Sample: VP/VLBW Survivors to DischargeEligible Sample in AdulthoodAssessed Sample in AdulthoodSample with data on ASR Adaptive Functioning ScaleaVP/VLBW Dropout and Missing Data, %VP/VLBW With NSI Dropout and Missing Data, %Term-Born Control Adults With Data on ASR Adaptive Functioning ScaleaComments
EPICure (39) The United Kingdom and Ireland National 1995 19 <26 wk of gestation 315 306 129 116 62.1 71.1 62 (recruited at ages 6 or 11) Birth wt and gestational age not available for control participants. Dropout associated with disability and social disadvantage 
ESTER (40,41) Finland Regional 1985-89 23 <37 wk of gestation NA NA 77 69 60.2b NA 330 (recruited at infancy) SES measured by parent education collected retrospectively; data on IVH and BPD not available at the time of data analysis. Dropout associated with male sex and physical disabilities 
HESVA
(42,43) 
Finland Regional 1978–1985 25 Birth wt ≤1500 g 334 254 165 112 55.9 59.1 104 (recruited in adulthood, no SGA term-born control participants) Maternal education measured retrospectively. Overrepresentation of healthier participants 
NTNU (8,9) Norway Regional 1986–1988 26 Birth wt ≤1500 g 86 82 62 58 29.2 45.5 88 (recruited at infancy but no SGA term-born control participants) Maternal education measured at the 14-y assessment. Dropout associated with male sex 
AYLS (44) Finland Regional 1985–1986 26 Admission for neonatal care within 10 d after birthc 108 68 35 30 55.9 66.7 316 (recruited at infancy) Dropout associated with lower parental education, male sex, and lower cognitive function 
CohortCountryScaleBirth YearMean Age of Assessment, yOriginal Index Eligibility CriteriaInitial Sample: VP/VLBW Survivors to DischargeEligible Sample in AdulthoodAssessed Sample in AdulthoodSample with data on ASR Adaptive Functioning ScaleaVP/VLBW Dropout and Missing Data, %VP/VLBW With NSI Dropout and Missing Data, %Term-Born Control Adults With Data on ASR Adaptive Functioning ScaleaComments
EPICure (39) The United Kingdom and Ireland National 1995 19 <26 wk of gestation 315 306 129 116 62.1 71.1 62 (recruited at ages 6 or 11) Birth wt and gestational age not available for control participants. Dropout associated with disability and social disadvantage 
ESTER (40,41) Finland Regional 1985-89 23 <37 wk of gestation NA NA 77 69 60.2b NA 330 (recruited at infancy) SES measured by parent education collected retrospectively; data on IVH and BPD not available at the time of data analysis. Dropout associated with male sex and physical disabilities 
HESVA
(42,43) 
Finland Regional 1978–1985 25 Birth wt ≤1500 g 334 254 165 112 55.9 59.1 104 (recruited in adulthood, no SGA term-born control participants) Maternal education measured retrospectively. Overrepresentation of healthier participants 
NTNU (8,9) Norway Regional 1986–1988 26 Birth wt ≤1500 g 86 82 62 58 29.2 45.5 88 (recruited at infancy but no SGA term-born control participants) Maternal education measured at the 14-y assessment. Dropout associated with male sex 
AYLS (44) Finland Regional 1985–1986 26 Admission for neonatal care within 10 d after birthc 108 68 35 30 55.9 66.7 316 (recruited at infancy) Dropout associated with lower parental education, male sex, and lower cognitive function 

NA, not applicable.

a

Participants with data on any of the 5 Adaptive Functioning scales (education, job, spouse or partner, friends, and family).

b

Extracted from the published article as an estimate of dropout of VP and/or VLBW participants (34). Gestational age and birth weight available only for participants assessed in adulthood, so it was impossible to calculate numbers of VP and/or VLBW survivors to discharge and the eligible sample in adulthood, as well as drop-out rates.

c

AYLS included infants who had been admitted to the neonatal ward or had been transferred to the NICU within 10 d after birth, so the original index group ranged from critically ill preterm children to term-born children with minimal complications and inpatient observation times.

Although 468 VP and/or VLBW and 1056 term-born adults met the eligibility criteria for the analysis, outcome data were only available for 385 VP and/or VLBW and 900 term-born adults. Among those assessed, there were no significant differences between VP and/or VLBW adults and those in the control group in the proportions of participants with data on all scales, except for the job scale in EPICure and ESTER (Supplemental Table 5). Missing data could have resulted from participant failure to complete the questionnaire, participant failure to provide individual responses required to calculate the corresponding scale, or nonapplicability of the scales if respondents were not living with a partner, did not have a job, or were not in education during the preceding 6 months. Because of variations in ages at assessment, younger participants were less likely to live with a partner or be employed, whereas older participants were less likely to be in education. For instance, in EPICure (age 19 years), only 1.6% (2 of 129) of assessed VP and/or VLBW participants lived with a partner and 6.2% (4 of 65) of adults in the control group; only 43.4% of VP and/or VLBW participants were employed and 66.2% of those in the control group (Supplemental Table 5). In contrast to EPICure, in NTNU (age 26 years), 24.2% of VP and/or VLBW participants were in education and 37.8% of adults in the control group.

In Table 2, we provide a summary of adult outcomes, NSI, and demographic and perinatal data for VP- and/or VLBW- and term-born adults in each cohort. Results were similar between 1-stage and 2-stage approaches (Fig 2) (Table 3) (Supplemental Table 6). We only report the 1-stage results in the main text. VP and/or VLBW participants reported significantly lower ratings on the friends scale compared with those in the control group (Δ −0.37, 95% CI: −0.61 to −0.13). This difference remained significant after adjusting for sex, age at assessment, and SES (Δ −0.39, 95% CI: −0.63 to −0.15). The sex and group interaction was insignificant (P = .782). Effect size was slightly reduced after excluding participants with childhood NSI (Δ −0.34, 95% CI: −0.61 to −0.07). There were no significant differences in the other scales (Fig 2).

FIGURE 2

A comparison of self-reported social functioning in adulthood between VP and/or VLBW participants and term-born adults in the control group. Notes: EPICure was not included for the spouse or partner scale because of small numbers. Effect size is measured by mean difference and 95% CIs for each scale. In the 2-stage approach, the size of the squares and the diamonds are proportional to the weight assigned to the relative effect sizes; heterogeneity was quantified by I2 and τ2. One-stage and 2-stage overall effects are represented by diamonds. Between-study heterogeneity between the 2-stage and 1-stage approaches were compared by using τ2.

FIGURE 2

A comparison of self-reported social functioning in adulthood between VP and/or VLBW participants and term-born adults in the control group. Notes: EPICure was not included for the spouse or partner scale because of small numbers. Effect size is measured by mean difference and 95% CIs for each scale. In the 2-stage approach, the size of the squares and the diamonds are proportional to the weight assigned to the relative effect sizes; heterogeneity was quantified by I2 and τ2. One-stage and 2-stage overall effects are represented by diamonds. Between-study heterogeneity between the 2-stage and 1-stage approaches were compared by using τ2.

Close modal
TABLE 2

Characteristics of VP and/or VLBW Adults and Term-Born Control Adults

EPICureESTERHESVANTNUAYLS
VP/VLBW Adults, n = 116Control Adults, n = 62PVP/VLBW Adults, n = 69Control Adults, n = 330PVP/VLBW Adults, n = 112Control Adults, n = 104PVP/VLBW Adults, n = 58Control Adults, n = 88PVP/VLBW Adults, n = 30Control Adults, n = 316P
Adult outcomesa                
 Family z score −0.18 (1.22) (n = 115) −0.00 (1.00) (n = 62) .330 0.04 (0.98) (n = 69) −0.01 (1.00) (n = 328) .711 −0.09 (0.94) (n = 112) 0.00 (1.00) (n = 104) .494 −0.15 (1.02) (n = 57) −0.00 (1.00) (n = 88) .394 0.06 (0.86) (n = 30) 0.00 (1.00) (n = 311) .760 
 Friends z score −0.79 (1.23) (n = 116) −0.00 (1.00) (n = 62) <.001 −0.20 (1.05) (n = 68) 0.03 (0.99) (n = 330) .087 −0.40 (1.03) (n = 108) 0.00 (1.00) (n = 101) .005 −0.39 (1.30) (n = 58) 0.00 (1.00) (n = 88) .042 0.00 (1.09) (n = 30) 0.00 (1.00) (n = 315) .972 
 Spouse/partner z score 0.86 (0.00) (n = 2) 0.00 (1.00) (n = 4) .315 0.20 (0.78) (n = 42) −0.01 (0.97) (n = 227) .186 0.04 (1.26) (n = 60) −0.00 (1.00) (n = 63) .837 0.12 (0.80) (n = 34) 0.00 (1.00) (n = 48) .577 −0.32 (1.01) (n = 21) 0.00 (1.00) (n = 229) .169 
 Job z score −0.10 (0.83) (n = 56) −0.00 (1.00) (n = 43) .594 0.08 (0.95) (n = 45) 0.01 (0.94) (n = 250) .641 −0.05 (0.87) (n = 91) 0.00 (1.00) (n = 93) .701 −0.54 (1.36) (n = 48) 0.00 (1.00) (n = 79) .012 0.11 (0.63) (n = 30) 0.00 (1.00) (n = 299) .524 
 Education z score 0.16 (0.89) (n = 89) 0.00 (1.00) (n = 51) .345 −0.12 (0.99) (n = 42) 0.02 (0.97) (n = 196) .378 −0.01 (1.15) (n = 57) −0.00 (1.00) (n = 64) .954 −0.38 (1.03) (n = 15) 0.00 (1.00) (n = 34) .230 0.04 (0.88) (n = 26) −0.00 (1.00) (n = 297) .778 
Demographic data                
Age at assessment in adulthood, y 19.3 (0.6) (n = 116) 19.2 (0.5) (n = 62) .323 23.1 (1.4) (n = 69) 23.5 (1.2) (n = 330) .003 24.6 (2.1) (n = 112) 24.6 (2.2) (n = 104) .841 26.3 (0.7) (n = 58) 26.5 (0.5) (n = 88) .093 25.8 (0.6) (n = 30) 25.5 (0.6) (n = 316) .009 
Men 45.7 (53 of 116) 37.1 (23 of 62) .269 42.0 (29 of 69) 46.5 (153 of 329) .497 43.8 (49 of 112) 43.3 (45 of 104) .943 53.4 (31 of 58) 43.2 (38 of 88) .224 56.7 (17 of 30) 57.3 (181 of 316) .141 
SES                
 Low 20.0 (22 of 110) 4.8 (3 of 62) <.001 5.8 (4 of 69) 6.1 (20 of 327) .963 15.5 (17 of 110) 14.4 (15 of 104) .498 6.4 (3 of 47) 6.8 (5 of 74) .943 26.7 (8 of 30) 18.5 (58 of 314) .550 
 Medium 76.4 (84 of 110) 75.8 (47 of 62)  62.3 (43 of 69) 60.6 (198 of 327)  61.8 (68 of 110) 55.8 (58 of 104)  53.2 (25 of 47) 50.0 (37 of 74)  30.0 (9 of 30) 32.5 (102 of 314)  
 High 3.6 (4 of 110) 19.4 (12 of 62)  31.9 (22 of 69) 33.3 (109 of 327)  22.7 (25 of 110) 29.8 (31 of 104)  40.4 (19 of 47) 43.2 (32 of 74)  43.3 (13 of 30) 49.0 (154 of 314)  
Childhood NSI,b 20.0 (22 of 110) 0.0 (0 of 51) .001 10.1 (7 of 69) 0.3 (1 of 330) <.001 8.3 (9 of 109) 1.0 (1 of 103) .012 16.2 (6 of 37) 0.0 (0 of 63) .001 21.4 (6 of 28) 1.1 (3 of 264) <.001 
Perinatal data                
 Gestational age, wk 24.5 (0.7) (n = 116) — — 30.6 (2.0) (n = 69) 40.1 (1.2) (n = 330) <.001 29.3 (2.4) (n = 112) 40.1 (1.1) (n = 104) <.001 28.8 (2.6) (n = 58) 39.8 (1.2) (n = 88) <.001 29.2 (2.4) (n = 30) 39.7 (1.2) (n = 316) <.001 
 Birth wt, g 745 (122) (n = 116) — — 1478 (406) (n = 69) 3572 (491) (n = 329) <.001 1128 (220) (n = 112) 3611 (492) (n = 104) <.001 1193 (247) (n = 58) 3702 (451) (n = 88) <.001 1317 (316) (n = 30) 3622 (470) (n = 316) <.001 
 Birth weight z score 0.23 (0.81) (n = 116) — — 0.07 (1.18) (n = 69) 0.09 (0.94) (n = 329) .935 −0.43 (1.01) (n = 112) 0.15 (1.00) (n = 104) <.001 −0.13 (1.04) (n = 58) 0.30 (0.81) (n = 88) .006 0.03 (1.02) (n = 30) 0.17 (0.86) (n = 316) .396 
 Multiple birth 33.9 (39 of 115) — — 30.4 (21 of 69) 1.2 (4 of 330) <.001 15.2 (17 of 112) 0.0 (0 of 104) <.001 15.5 (9 of 58) 0.0 (0 of 88) <.001 13.3 (4 of 30) 1.3 (4 of 316) <.001 
 IVH Grades 3–4 12.2 (14 of 115) — — — — — 4.8 (4 of 83) — — 2.1 (1 of 47) — — 6.7 (2 of 30) 0.0 (0 of 316) <.001 
 IVH 67.0 (77 of 115) — — — — — 16.9 (14 of 83) — — 12.8 (6 of 47) — — 20.0 (6 of 30) 0.0 (0 of 316) <.001 
 BPD 72.4 (84 of 116) — — — — — 24.1 (26 of 108) — — 23.2 (13 of 56) 0.0 (0 of 88) <.001 3.3 (1 of 30) 0.0 (0 of 316) .001 
EPICureESTERHESVANTNUAYLS
VP/VLBW Adults, n = 116Control Adults, n = 62PVP/VLBW Adults, n = 69Control Adults, n = 330PVP/VLBW Adults, n = 112Control Adults, n = 104PVP/VLBW Adults, n = 58Control Adults, n = 88PVP/VLBW Adults, n = 30Control Adults, n = 316P
Adult outcomesa                
 Family z score −0.18 (1.22) (n = 115) −0.00 (1.00) (n = 62) .330 0.04 (0.98) (n = 69) −0.01 (1.00) (n = 328) .711 −0.09 (0.94) (n = 112) 0.00 (1.00) (n = 104) .494 −0.15 (1.02) (n = 57) −0.00 (1.00) (n = 88) .394 0.06 (0.86) (n = 30) 0.00 (1.00) (n = 311) .760 
 Friends z score −0.79 (1.23) (n = 116) −0.00 (1.00) (n = 62) <.001 −0.20 (1.05) (n = 68) 0.03 (0.99) (n = 330) .087 −0.40 (1.03) (n = 108) 0.00 (1.00) (n = 101) .005 −0.39 (1.30) (n = 58) 0.00 (1.00) (n = 88) .042 0.00 (1.09) (n = 30) 0.00 (1.00) (n = 315) .972 
 Spouse/partner z score 0.86 (0.00) (n = 2) 0.00 (1.00) (n = 4) .315 0.20 (0.78) (n = 42) −0.01 (0.97) (n = 227) .186 0.04 (1.26) (n = 60) −0.00 (1.00) (n = 63) .837 0.12 (0.80) (n = 34) 0.00 (1.00) (n = 48) .577 −0.32 (1.01) (n = 21) 0.00 (1.00) (n = 229) .169 
 Job z score −0.10 (0.83) (n = 56) −0.00 (1.00) (n = 43) .594 0.08 (0.95) (n = 45) 0.01 (0.94) (n = 250) .641 −0.05 (0.87) (n = 91) 0.00 (1.00) (n = 93) .701 −0.54 (1.36) (n = 48) 0.00 (1.00) (n = 79) .012 0.11 (0.63) (n = 30) 0.00 (1.00) (n = 299) .524 
 Education z score 0.16 (0.89) (n = 89) 0.00 (1.00) (n = 51) .345 −0.12 (0.99) (n = 42) 0.02 (0.97) (n = 196) .378 −0.01 (1.15) (n = 57) −0.00 (1.00) (n = 64) .954 −0.38 (1.03) (n = 15) 0.00 (1.00) (n = 34) .230 0.04 (0.88) (n = 26) −0.00 (1.00) (n = 297) .778 
Demographic data                
Age at assessment in adulthood, y 19.3 (0.6) (n = 116) 19.2 (0.5) (n = 62) .323 23.1 (1.4) (n = 69) 23.5 (1.2) (n = 330) .003 24.6 (2.1) (n = 112) 24.6 (2.2) (n = 104) .841 26.3 (0.7) (n = 58) 26.5 (0.5) (n = 88) .093 25.8 (0.6) (n = 30) 25.5 (0.6) (n = 316) .009 
Men 45.7 (53 of 116) 37.1 (23 of 62) .269 42.0 (29 of 69) 46.5 (153 of 329) .497 43.8 (49 of 112) 43.3 (45 of 104) .943 53.4 (31 of 58) 43.2 (38 of 88) .224 56.7 (17 of 30) 57.3 (181 of 316) .141 
SES                
 Low 20.0 (22 of 110) 4.8 (3 of 62) <.001 5.8 (4 of 69) 6.1 (20 of 327) .963 15.5 (17 of 110) 14.4 (15 of 104) .498 6.4 (3 of 47) 6.8 (5 of 74) .943 26.7 (8 of 30) 18.5 (58 of 314) .550 
 Medium 76.4 (84 of 110) 75.8 (47 of 62)  62.3 (43 of 69) 60.6 (198 of 327)  61.8 (68 of 110) 55.8 (58 of 104)  53.2 (25 of 47) 50.0 (37 of 74)  30.0 (9 of 30) 32.5 (102 of 314)  
 High 3.6 (4 of 110) 19.4 (12 of 62)  31.9 (22 of 69) 33.3 (109 of 327)  22.7 (25 of 110) 29.8 (31 of 104)  40.4 (19 of 47) 43.2 (32 of 74)  43.3 (13 of 30) 49.0 (154 of 314)  
Childhood NSI,b 20.0 (22 of 110) 0.0 (0 of 51) .001 10.1 (7 of 69) 0.3 (1 of 330) <.001 8.3 (9 of 109) 1.0 (1 of 103) .012 16.2 (6 of 37) 0.0 (0 of 63) .001 21.4 (6 of 28) 1.1 (3 of 264) <.001 
Perinatal data                
 Gestational age, wk 24.5 (0.7) (n = 116) — — 30.6 (2.0) (n = 69) 40.1 (1.2) (n = 330) <.001 29.3 (2.4) (n = 112) 40.1 (1.1) (n = 104) <.001 28.8 (2.6) (n = 58) 39.8 (1.2) (n = 88) <.001 29.2 (2.4) (n = 30) 39.7 (1.2) (n = 316) <.001 
 Birth wt, g 745 (122) (n = 116) — — 1478 (406) (n = 69) 3572 (491) (n = 329) <.001 1128 (220) (n = 112) 3611 (492) (n = 104) <.001 1193 (247) (n = 58) 3702 (451) (n = 88) <.001 1317 (316) (n = 30) 3622 (470) (n = 316) <.001 
 Birth weight z score 0.23 (0.81) (n = 116) — — 0.07 (1.18) (n = 69) 0.09 (0.94) (n = 329) .935 −0.43 (1.01) (n = 112) 0.15 (1.00) (n = 104) <.001 −0.13 (1.04) (n = 58) 0.30 (0.81) (n = 88) .006 0.03 (1.02) (n = 30) 0.17 (0.86) (n = 316) .396 
 Multiple birth 33.9 (39 of 115) — — 30.4 (21 of 69) 1.2 (4 of 330) <.001 15.2 (17 of 112) 0.0 (0 of 104) <.001 15.5 (9 of 58) 0.0 (0 of 88) <.001 13.3 (4 of 30) 1.3 (4 of 316) <.001 
 IVH Grades 3–4 12.2 (14 of 115) — — — — — 4.8 (4 of 83) — — 2.1 (1 of 47) — — 6.7 (2 of 30) 0.0 (0 of 316) <.001 
 IVH 67.0 (77 of 115) — — — — — 16.9 (14 of 83) — — 12.8 (6 of 47) — — 20.0 (6 of 30) 0.0 (0 of 316) <.001 
 BPD 72.4 (84 of 116) — — — — — 24.1 (26 of 108) — — 23.2 (13 of 56) 0.0 (0 of 88) <.001 3.3 (1 of 30) 0.0 (0 of 316) .001 

—, not applicable.

a

Mean (SD) for continuous variables and % (n of n) for categorical variables.

a

Raw-sum score was converted to z score by using mean and SD of individuals in the control group for each scale within each cohort.

b

NSI was defined as ≥1 of the following: visual impairment (blind in 1 or both eyes), hearing impairment (not corrected by hearing aids), cerebral palsy (nonambulatory [ie, use of wheelchair or bed ridden]), or cognitive impairment in childhood.

TABLE 3

Univariate Analysis of Factors Associated With Social Functioning Within VP and/or VLBW Participants: Linear Mixed-Effects Models

Family, Effect (95% CI)bFriends, Effect (95% CI)bSpouse/Partnera, Effect (95% CI)bJob, Effect (95% CI)bEducation, Effect (95% CI)b
Age at assessment 0.00 (−0.08 to 0.08) (n = 383) 0.01 (−0.07 to 0.10) (n = 380) 0.06 (−0.05 to 0.17) (n = 157) −0.06 (−0.17 to 0.05) (n = 270) −0.01 (−0.21 to 0.19) (n = 229) 
Sex (reference = male) −0.05 (−0.28 to 0.17) (n = 383) −0.06 (−0.29 to 0.18) (n = 380) −0.14 (−0.65 to 0.36) (n = 157) 0.09 (−0.36 to 0.54) (n = 270) −0.12 (−0.38 to 0.14) (n = 229) 
SES      
 Medium versus low 0.27 (−0.05 to 0.59) (n = 364) 0.27 (−0.08 to 0.62) (n = 361) 0.42 (−0.12 to 0.97) (n = 151) 0.09 (−0.27 to 0.45) (n = 256) 0.16 (−0.28 to 0.60) (n = 222) 
 High versus low 0.22 (−0.15 to 0.60) (n = 364) 0.29 (−0.13 to 0.70) (n = 361) 0.69 (0.09 to 1.28) (n = 151) −0.01 (−0.41 to 0.39) (n = 256) 0.11 (−0.47 to 0.69) (n = 222) 
NSI in childhood (reference = no) −0.24 (−0.80 to 0.31) (n = 351) −0.48 (−1.02 to 0.06) (n = 348) −0.40 (−1.01 to 0.22) (n = 144) −0.14 (−0.53 to 0.24) (n = 246) −0.37 (−0.74 to 0.00) (n = 217) 
Gestational age, wk 0.04 (−0.03 to 0.11) (n = 383) 0.06 (−0.02 to 0.13) (n = 380) −0.03 (−0.10 to 0.05) (n = 157) −0.03 (−0.11 to 0.06) (n = 270) −0.03 (−0.10 to 0.04) (n = 229) 
Birth weight z score −0.04 (−0.15 to 0.06) (n = 383) 0.01 (−0.14 to 0.16) (n = 380) 0.08 (−0.07 to 0.23) (n = 157) 0.03 (−0.10 to 0.16) (n = 270) 0.05 (−0.09 to 0.19) (n = 229) 
Multiple birth (reference = singleton) 0.29 (0.00 to 0.58) (n = 382) 0.05 (−0.33 to 0.43) (n = 379) 0.31 (−0.31 to 0.92) (n = 157) 0.08 (−0.21 to 0.38) (n = 269) 0.15 (−0.29 to 0.59) (n = 228) 
IVH grade 3–4 (reference = no IVH or IVH grade 1–2)c −0.70 (−1.19 to −0.21) (n = 273) −0.39 (−0.93 to 0.14) (n = 271) −1.27 (−3.50 to 0.96) (n = 91) 0.19 (−0.41 to 0.78) (n = 194) 0.20 (−0.34 to 0.74) (n = 178) 
IVH (reference = no)c −0.12 (−0.43 to 0.19) (n = 273) 0.24 (−0.11 to 0.59) (n = 271) −0.78 (−1.52 to −0.03) (n = 91) 0.07 (−0.42 to 0.56) (n = 194) 0.14 (−0.44 to 0.72) (n = 178) 
BPD (reference = no)c −0.18 (−0.75 to 0.40) (n = 308) −0.46 (−0.77 to −0.15) (n = 306)d 0.15 (−0.40 to 0.71) (n = 112) −0.16 (−0.47 to 0.14) (n = 220) −0.23 (−0.58 to 0.11) (n = 185) 
Family, Effect (95% CI)bFriends, Effect (95% CI)bSpouse/Partnera, Effect (95% CI)bJob, Effect (95% CI)bEducation, Effect (95% CI)b
Age at assessment 0.00 (−0.08 to 0.08) (n = 383) 0.01 (−0.07 to 0.10) (n = 380) 0.06 (−0.05 to 0.17) (n = 157) −0.06 (−0.17 to 0.05) (n = 270) −0.01 (−0.21 to 0.19) (n = 229) 
Sex (reference = male) −0.05 (−0.28 to 0.17) (n = 383) −0.06 (−0.29 to 0.18) (n = 380) −0.14 (−0.65 to 0.36) (n = 157) 0.09 (−0.36 to 0.54) (n = 270) −0.12 (−0.38 to 0.14) (n = 229) 
SES      
 Medium versus low 0.27 (−0.05 to 0.59) (n = 364) 0.27 (−0.08 to 0.62) (n = 361) 0.42 (−0.12 to 0.97) (n = 151) 0.09 (−0.27 to 0.45) (n = 256) 0.16 (−0.28 to 0.60) (n = 222) 
 High versus low 0.22 (−0.15 to 0.60) (n = 364) 0.29 (−0.13 to 0.70) (n = 361) 0.69 (0.09 to 1.28) (n = 151) −0.01 (−0.41 to 0.39) (n = 256) 0.11 (−0.47 to 0.69) (n = 222) 
NSI in childhood (reference = no) −0.24 (−0.80 to 0.31) (n = 351) −0.48 (−1.02 to 0.06) (n = 348) −0.40 (−1.01 to 0.22) (n = 144) −0.14 (−0.53 to 0.24) (n = 246) −0.37 (−0.74 to 0.00) (n = 217) 
Gestational age, wk 0.04 (−0.03 to 0.11) (n = 383) 0.06 (−0.02 to 0.13) (n = 380) −0.03 (−0.10 to 0.05) (n = 157) −0.03 (−0.11 to 0.06) (n = 270) −0.03 (−0.10 to 0.04) (n = 229) 
Birth weight z score −0.04 (−0.15 to 0.06) (n = 383) 0.01 (−0.14 to 0.16) (n = 380) 0.08 (−0.07 to 0.23) (n = 157) 0.03 (−0.10 to 0.16) (n = 270) 0.05 (−0.09 to 0.19) (n = 229) 
Multiple birth (reference = singleton) 0.29 (0.00 to 0.58) (n = 382) 0.05 (−0.33 to 0.43) (n = 379) 0.31 (−0.31 to 0.92) (n = 157) 0.08 (−0.21 to 0.38) (n = 269) 0.15 (−0.29 to 0.59) (n = 228) 
IVH grade 3–4 (reference = no IVH or IVH grade 1–2)c −0.70 (−1.19 to −0.21) (n = 273) −0.39 (−0.93 to 0.14) (n = 271) −1.27 (−3.50 to 0.96) (n = 91) 0.19 (−0.41 to 0.78) (n = 194) 0.20 (−0.34 to 0.74) (n = 178) 
IVH (reference = no)c −0.12 (−0.43 to 0.19) (n = 273) 0.24 (−0.11 to 0.59) (n = 271) −0.78 (−1.52 to −0.03) (n = 91) 0.07 (−0.42 to 0.56) (n = 194) 0.14 (−0.44 to 0.72) (n = 178) 
BPD (reference = no)c −0.18 (−0.75 to 0.40) (n = 308) −0.46 (−0.77 to −0.15) (n = 306)d 0.15 (−0.40 to 0.71) (n = 112) −0.16 (−0.47 to 0.14) (n = 220) −0.23 (−0.58 to 0.11) (n = 185) 

Linear mixed-effects models were based on the REML estimator, accounting for clustering. Variables of interest were preferably modeled as a random effect.

a

EPICure was not included for the spouse or partner scale because of small numbers. For IVH grades 3 to 4, data were insufficient for a random-effects model, so a fixed-effects model was used.

b

For categorical variables, this refers to mean difference; for continuous variables (age at assessment, birth weight, and gestational age), this refers to the average change in the outcome per unit increase in the variable.

c

No data in ESTER.

d

Remains significant after correcting for multiple comparisons by using the Benjamini-Hochberg procedure, with a false discovery rate of 0.05.

Within VP and/or VLBW participants, BPD diagnosis was significantly associated with lower ratings on the friends scale; IVH and low SES associated with lower ratings on the spouse or partner scale; and singleton birth and IVH grades 3 to 4 associated with lower ratings on the family scale (Table 3). After correcting for multiple comparisons, only the relationship between BPD and the friends scale remained significant. We further analyzed the difference on the friends scale between VP and/or VLBW and control groups after excluding participants with BPD. It remained significant but with a smaller effect size (Δ −0.25, 95% CI: −0.39 to −0.10).

This IPD meta-analysis investigated VP and/or VLBW adults’ perceptions of 5 different aspects of social functioning. We found that VP and/or VLBW adults perceived their relations with partner and family and their experiences of work and education (eg, satisfaction, ability to finish tasks, and relationship with classmates and/or colleagues) as similar to those in the control group. However, VP and/or VLBW adults reported significantly lower ratings for their relationships with friends, with the lowest reported by those with BPD.

The findings of similar relations with partners and work and educational experiences between VP and/or VLBW adults and control adults appear at odds with studies in which researchers use factual information that revealed a decreased likelihood of having a partner, lower educational attainment, and higher rates of unemployment in preterm adults.5  In our study, we focus on individuals’ perceptions of relationships and experiences and provide complementary information on adult social lives in different roles. Our results reassuringly correspond to previous research on the quality of romantic relationship5  and further extend to other areas of self-perceived social functioning (eg, work and education) in adulthood. Among VP and/or VLBW adults who had a partner or a job or were in education, the quality of the relationship and the perception of work and educational experiences were comparable to those of adults in the control group. Dissimilarity in findings between factual data and individual perceptions might arise because, opposed to objective circumstances, people tend to perceive their abilities and attributes more favorably compared with an average peer in one’s school or workplace: a phenomenon termed as “the better than average effect” in the field of social comparison.45,46 

Consistent with existing studies,713  we found similar family relationships between the 2 groups. In contrast, VP and/or VLBW adults scored their relationship with friends significantly lower (scores based on the number of close friends, frequency of contacts and/or visits, and getting along with close friends). The difference remained significant after excluding participants with childhood NSI or those with BPD. Furthermore, we did not find a difference on the friends scale between VP and/or VLBW female adults and term-born female adults, as previously suggested.10  In previous studies, researchers have reported VLBW participants having fewer friends in adulthood.14  Our finding provides valuable insights into the understanding of friendship and peer relationships in VP and/or VLBW adults. It indicates that friendship characteristics (eg, fewer friends, less satisfaction, and less time spent with friends) and peer social difficulties (eg, measured by the Strengths and Difficulties Questionnaire) reported in childhood4751  may persist into adulthood. We speculate that less interaction with friends might be explained by mental health problems, personality or autistic traits, behavioral characteristics, or motor difficulties among VP and/or VLBW adults. VP/VLBW adults have been found to be less socially engaged, more easily worried and/or anxious, less confident in social relationships, having poorer motor skills, experiencing more challenges in social interaction and communication (autistic features), and exhibiting less frequent risk-taking behaviors.5256  All these characteristics might be interrelated and share common causes in early life.

The mechanisms underlying the link between VP birth and adulthood friendship are likely to be complex. One is linked to biological risk: VP birth leads to biological changes affecting long-term social functioning. Research has revealed that changes in the brain development in premature infants, such as alterations in amygdala functional connectivity and regional cerebral development, persist into adulthood.5762  The amygdala, together with a broad array of other brain regions, plays a central role in the social life of both VP and term individuals.63  Amygdala volume and functional connectivity with cortical regions correlate with measures of social functioning,57,64,65  including social network size. Furthermore, early alterations in regional cerebral development in VP infants are related to social development by school age.60 

The association between VP birth and friendship may also be explained by poorer motor and cognitive abilities after VP birth. Evidence supports the associations of higher motor and cognitive abilities with having more friends and spending more time with them in childhood and as adults.48,66  Indeed, childhood NSI partly explained the difference in scores on the friends scale; we found a reduced effect size after excluding VP and/or VLBW participants with NSI. On average, VP and/or VLBW individuals have more motor problems that often persist into adulthood5355 ,67  and may thus participate less in physical activity,68  further relating to more peer problems.69,70  Furthermore, cognitive ability might be more influential on social competence than is physical ability.47  Harrison suggested that VLBW children have cognitive and behavioral deficits that isolate them from both their peers and their peers’ risk-taking behavior in the absence of NSI.71  Cognitive ability is important for the development of social skills, such as emotion regulation, ability to read cues, and interact through gestures and body language.72  Relatedly, people may be more accepting of social interactions with peers with physical disabilities than those with cognitive disabilities.47 

Additionally, the lower scores on the friends scale in VP and/or VLBW adults could be due to alternations in hypothalamus-pituitary-adrenal axis stress-regulating system. There is some evidence suggesting a poorer response to acute (psychosocial) stress in both VP children and adults.73  In other studies, researchers have further linked blunted hypothalamus-pituitary-adrenal axis responses to stress and lower cortisol levels with some personality traits (eg, less openness),74  which might subsequently affect how VP and/or VLBW adults deal with social relationships with friends.

The other mechanism relates to environmental influences: social experiences play an important role in forming and maintaining friendships, such as early parent-infant relationship, parenting, and childhood peer relationships or bullying experiences. There is evidence that higher gestational age predicted better quality of early parent-infant relationships,50  which subsequently predicted higher friendship scores50  and better peer relationships in childhood.66  Research also reveals that mothers of VP and/or VLBW children may be more controlling.75  This may in turn lead to social isolation, with fewer opportunities for risk-taking behaviors as adults.15,42  Moreover, VP and/or VLBW children are more likely to be excluded and bullied at school,47,7678  which may adversely affect adult peer relationships in both preterm and general population samples.77,79  Finally, the biological and environmental pathways may individually affect or interact with each other, with additive or reinforced effects in influencing adult outcomes. Recent evidence also suggests environmental factors (eg, bullying) as mediators between VP birth and adult mental health outcomes.80 

In our study, we also identify that VP and/or VLBW adults with BPD were at greatest risk of having lower scores on the friends scale, suggesting persisting nonrespiratory effects of BPD into adulthood. To our knowledge, this finding has not been previously reported. The BPD-friendship association may be mediated through the associated reduced lung function and poorer motor function and cognitive development through childhood and into adulthood.8183  Because of these limitations, VP and/or VLBW adults with BPD may have more problems with usual activities, such as hobbies, sports, and doing things with friends,4  eventually leading to fewer friends or poorer relations. We also showed that IVH and low SES were significantly associated with lower scores on the spouse or partner scale and that singleton birth and IVH grades 3 to 4 were related to lower scores on the family scale, but these became statistically nonsignificant after correcting for multiple comparisons.

The strengths of this study include the large sample size, combining IPD from 5 European cohorts; the availability of comprehensive perinatal and childhood data, allowing us to control for confounders and explore their roles in relation to social functioning; the harmonization of IPD by using unified criteria to control for between-study heterogeneity; and the exploration of social functioning based on the personal view of life experiences measured by using the same self-report scale. Several limitations should be acknowledged. The first is differences in eligibility criteria among cohorts, such as the stricter inclusion criterion of <26 weeks' gestation in EPICure. Secondly, potential bias could be attributed to dropout (due to disability, social disadvantage, etc) and missing data, resulting in overrepresentation of healthier participants. Our differences in friendship between VP and/or VLBW and control groups may therefore underestimate the true differences. However, this is likely to be minimal because NSI and SES were unrelated to friendship in our study. Thirdly, the finding of the positive association between BPD and friendship needs replication because of variations in the definition of BPD. Fourthly, although self-report questionnaires offer participants the opportunity to describe their own experiences, they are susceptible to report bias associated with social desirability, applicable to both groups. Lastly, survival for infants born at extremely low gestational ages (22–25 weeks) is continuously increasing. It is likely that only a small proportion of infants in this analysis were born at these gestational ages because all recruited participants were born before 1996. Therefore, caution should be taken when generalizing our research findings to surviving VP infants born in recent decades.

In this analysis, we highlight the value of including individual perceptions of social functioning and provide a comprehensive picture of adult social life. VP and/or VLBW adults rated their friendships, in terms of number, contact and/or visiting frequency, and satisfaction, lower than did term-born adults. In contrast, despite being exposed to considerable stress in early life, VP and/or VLBW adults perceived the quality of their family and partner relationships, as well as work and educational experiences, as comparable to those of controls. Considering the well-described, wide-ranging negative consequences after VP and/or VLBW birth, this finding is particularly reassuring. Early interventions in clinical, family, and school settings to enhance social skills and social inclusion in childhood may improve peer relationships of VP and/or VLBW individuals in adulthood.47,84 

Dr Ni had full access to anonymized individual participant data from all participating cohorts, conceptualized and conducted the data harmonization and the data analysis, drafted the initial manuscript, and revised the manuscript; Dr Mendonça contributed to the design of the study, bringing cohort data together, and the data harmonization; Ms Baumann and Mr Eves contributed to the data harmonization and the data analysis; Professor Wolke conceptualized and designed the study, obtained funding, had full access to anonymized individual participant data from all participating cohorts, and revised the manuscript; and all authors contributed to the data acquisition and the data interpretation, critically reviewed the manuscript for important intellectual content, approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

FUNDING: Funded by the European Union’s Horizon 2020 Research and Innovation Program (grant 733280). The funder did not participate in the work.

APIC

Adults born Preterm International Collaboration

ASR

Adult Self-Report

BPD

bronchopulmonary dysplasia

CI

confidence interval

IPD

individual participant data

ISCED

International Standard Classification of Education

IVH

intraventricular hemorrhage

NSI

neurosensory impairment

RECAP Preterm

Research on European Children and Adults born Preterm

REML

restricted maximum likelihood

SES

socioeconomic status

VLBW

very low birth weight

VP

very preterm

1
Wolke
D
,
Johnson
S
,
Mendonça
M
.
The life course consequences of very preterm birth
.
Annual Review of Developmental Psychology
.
2019
;
1
(
1
):
69
92
2
Saigal
S
,
Doyle
LW
.
An overview of mortality and sequelae of preterm birth from infancy to adulthood
.
Lancet
.
2008
;
371
(
9608
):
261
269
3
Hurst
JR
,
Beckmann
J
,
Ni
Y
, et al
.
Respiratory and cardiovascular outcomes in survivors of extremely preterm birth at 19 years
.
Am J Respir Crit Care Med
.
2020
;
202
(
3
):
422
432
4
Gough
A
,
Linden
M
,
Spence
D
,
Patterson
CC
,
Halliday
HL
,
McGarvey
LP
.
Impaired lung function and health status in adult survivors of bronchopulmonary dysplasia
.
Eur Respir J
.
2014
;
43
(
3
):
808
816
5
Mendonça
M
,
Bilgin
A
,
Wolke
D
.
Association of preterm birth and low birth weight with romantic partnership, sexual intercourse, and parenthood in adulthood: a systematic review and meta-analysis
.
JAMA Netw Open
.
2019
;
2
(
7
):
e196961
6
Bosc
M
.
Assessment of social functioning in depression
.
Compr Psychiatry
.
2000
;
41
(
1
):
63
69
7
Bilgin
A
,
Mendonca
M
,
Wolke
D
.
Preterm birth/low birth weight and markers reflective of wealth in adulthood: a meta-analysis
.
Pediatrics
.
2018
;
142
(
1
):
e20173625
8
Lund
LK
,
Vik
T
,
Lydersen
S
, et al
.
Mental health, quality of life and social relations in young adults born with low birth weight
.
Health Qual Life Outcomes
.
2012
;
10
:
146
9
Husby
IM
,
Stray
KM
,
Olsen
A
, et al
.
Long-term follow-up of mental health, health-related quality of life and associations with motor skills in young adults born preterm with very low birth weight
.
Health Qual Life Outcomes
.
2016
;
14
:
56
10
Hack
M
,
Youngstrom
EA
,
Cartar
L
, et al
.
Behavioral outcomes and evidence of psychopathology among very low birth weight infants at age 20 years
.
Pediatrics
.
2004
;
114
(
4
):
932
940
11
Saigal
S
.
Functional outcomes of very premature infants into adulthood
.
Semin Fetal Neonatal Med
.
2014
;
19
(
2
):
125
130
12
Saigal
S
,
Day
KL
,
Van Lieshout
RJ
,
Schmidt
LA
,
Morrison
KM
,
Boyle
MH
.
Health, wealth, social integration, and sexuality of extremely low-birth-weight prematurely born adults in the fourth decade of life
.
JAMA Pediatr
.
2016
;
170
(
7
):
678
686
13
Scharf
M
,
Cohen
T
.
Relatedness and individuation among young adults born preterm: the role of relationships with parents and death anxiety
.
J Adult Dev
.
2013
;
20
(
4
):
212
221
14
Darlow
BA
,
Horwood
LJ
,
Pere-Bracken
HM
,
Woodward
LJ
.
Psychosocial outcomes of young adults born very low birth weight
.
Pediatrics
.
2013
;
132
(
6
):
e1521
e1528
15
Hack
M
,
Flannery
DJ
,
Schluchter
M
,
Cartar
L
,
Borawski
E
,
Klein
N
.
Outcomes in young adulthood for very-low-birth-weight infants
.
N Engl J Med
.
2002
;
346
(
3
):
149
157
16
Saigal
S
,
Stoskopf
B
,
Streiner
D
, et al
.
Transition of extremely low-birth-weight infants from adolescence to young adulthood: comparison with normal birth-weight controls
.
JAMA
.
2006
;
295
(
6
):
667
675
17
Kajantie
E
,
Hovi
P
,
Räikkönen
K
, et al
.
Young adults with very low birth weight: leaving the parental home and sexual relationships--Helsinki Study of Very Low Birth Weight Adults
.
Pediatrics
.
2008
;
122
(
1
):
e62
e72
18
Twilhaar
ES
,
Wade
RM
,
de Kieviet
JF
,
van Goudoever
JB
,
van Elburg
RM
,
Oosterlaan
J
.
Cognitive outcomes of children born extremely or very preterm since the 1990s and associated risk factors: a meta-analysis and meta-regression
.
JAMA Pediatr
.
2018
;
172
(
4
):
361
367
19
Riley
RD
,
Lambert
PC
,
Abo-Zaid
G
.
Meta-analysis of individual participant data: rationale, conduct, and reporting
.
BMJ
.
2010
;
340
:
c221
20
Simmonds
M
,
Stewart
G
,
Stewart
L
.
A decade of individual participant data meta-analyses: A review of current practice
.
Contemp Clin Trials
.
2015
;
45
(
Pt A
):
76
83
21
Stewart
LA
,
Parmar
MK
.
Meta-analysis of the literature or of individual patient data: is there a difference?
Lancet
.
1993
;
341
(
8842
):
418
422
22
Burgess
S
,
White
IR
,
Resche-Rigon
M
,
Wood
AM
.
Combining multiple imputation and meta-analysis with individual participant data
.
Stat Med
.
2013
;
32
(
26
):
4499
4514
23
Stewart
LA
,
Clarke
M
,
Rovers
M
, et al;
PRISMA-IPD Development Group
.
Preferred Reporting Items for Systematic Review and Meta-Analyses of individual participant data: the PRISMA-IPD Statement
.
JAMA
.
2015
;
313
(
16
):
1657
1665
24
Wells
GA
,
Shea
B
,
O’Connell
D
, et al
.
The Newcastle-Ottawa Scale (NOS) for Assessing the Quality of Nonrandomised Studies in Meta-analyses
.
Ottowa, ON
:
The Ottowa Hospital Research Institute
;
2000
.
25
Fenton
TR
,
Kim
JH
.
A systematic review and meta-analysis to revise the Fenton growth chart for preterm infants
.
BMC Pediatr
.
2013
;
13
:
59
26
Jobe
AH
,
Bancalari
E
.
Bronchopulmonary dysplasia
.
Am J Respir Crit Care Med
.
2001
;
163
(
7
):
1723
1729
27
Papile
LA
,
Burstein
J
,
Burstein
R
,
Koffler
H
.
Incidence and evolution of subependymal and intraventricular hemorrhage: a study of infants with birth weights less than 1,500 gm
.
J Pediatr
.
1978
;
92
(
4
):
529
534
28
UNESCO Institute for Statistics
.
International Standard Classification of Education: ISCED 2011
.
Montreal
:
UNESCO Institute for Statistics
;
2012
29
Rescorla
LA
,
Achenbach
TM
,
Ivanova
MY
, et al
.
Problems and adaptive functioning reported by adults in 17 societies
.
Int Perspect Psychol
.
2016
;
5
(
2
):
91
109
30
Achenbach
TM
,
Rescorla
LA
.
Manual for the ASEBA Adult Forms & Profiles
.
Burlington, VT
:
University of Vermont: Research Center for Children, Youth, and Families
;
2003
31
Achenbach
TM
,
Rescorla
LA
.
Multicultural Supplement to the Manual for the ASEBA Adult Forms & Profiles
.
Burlington, VT
:
University of Vermont Research Center for Children, Youth, and Families
;
2015
.
32
Fisher
DJ
.
Two-stage individual participant data meta-analysis and generalized forest plots
.
Stata J
.
2015
;
15
(
2
):
369
396
33
Kontopantelis
E
,
Reeves
D
.
A short guide and a forest plot command (Ipdforest) for one-stage meta-analysis
.
Stata J
.
2013
;
13
(
3
):
574
587
34
Burke
DL
,
Ensor
J
,
Riley
RD
.
Meta-analysis using individual participant data: one-stage and two-stage approaches, and why they may differ
.
Stat Med
.
2017
;
36
(
5
):
855
875
35
Raudenbush
SW
.
Analyzing effect sizes: random-effects models
. In:
Cooper
H
,
Hedges
HV
,
Valentine
JC
, eds.
The Handbook of Research Synthesis and Meta-analysis
.
Manhattan, NY
:
Russell Sage Foundation
;
2009
:
295
316
36
Borenstein
M
,
Hedges
LV
,
Higgins
JP
,
Rothstein
HR
.
Introduction to meta-analysis
, 2nd ed.
Chichester, UK
:
John Wiley & Sons
;
2021
37
Deeks
JJ
,
Higgins
JP
,
Altman
DG
,
Group
CSM
.
Analysing data and undertaking meta‐analyses
. In:
Higgins
JPT
,
Thomas
J
,
Chandler
J
, et al, eds.
Cochrane Handbook for Systematic Reviews of Interventions
.
Chichester, UK
:
John Wiley & Sons
;
2019
:
241
284
38
McDonald
JH
.
Handbook of Biological Statistics
, 3rd ed.
Baltimore, MD
:
Sparky House Publishing
;
2014
39
Wood
NS
,
Marlow
N
,
Costeloe
K
,
Gibson
AT
,
Wilkinson
AR
.
Neurologic and developmental disability after extremely preterm birth. EPICure Study Group
.
N Engl J Med
.
2000
;
343
(
6
):
378
384
40
Männistö
T
,
Vääräsmäki
M
,
Sipola-Leppänen
M
, et al
.
Independent living and romantic relations among young adults born preterm
.
Pediatrics
.
2015
;
135
(
2
):
290
297
41
Sipola-Leppänen
M
,
Vääräsmäki
M
,
Tikanmäki
M
, et al
.
Cardiometabolic risk factors in young adults who were born preterm
.
Am J Epidemiol
.
2015
;
181
(
11
):
861
873
42
Pesonen
AK
,
Räikkönen
K
,
Heinonen
K
, et al
.
Personality of young adults born prematurely: the Helsinki study of very low birth weight adults
.
J Child Psychol Psychiatry
.
2008
;
49
(
6
):
609
617
43
Räikkönen
K
,
Pesonen
AK
,
Heinonen
K
, et al
.
Depression in young adults with very low birth weight: the Helsinki study of very low-birth-weight adults
.
Arch Gen Psychiatry
.
2008
;
65
(
3
):
290
296
44
Kumpulainen
SM
,
Heinonen
K
,
Salonen
MK
, et al
.
Childhood cognitive ability and body composition in adulthood
.
Nutr Diabetes
.
2016
;
6
(
8
):
e223
45
Zell
E
,
Alicke
MD
.
The local dominance effect in self-evaluation: evidence and explanations
.
Pers Soc Psychol Rev
.
2010
;
14
(
4
):
368
384
46
Zell
E
,
Strickhouser
JE
,
Sedikides
C
,
Alicke
MD
.
The better-than-average effect in comparative self-evaluation: a comprehensive review and meta-analysis
.
Psychol Bull
.
2020
;
146
(
2
):
118
149
47
Ritchie
K
,
Bora
S
,
Woodward
LJ
.
Social development of children born very preterm: a systematic review
.
Dev Med Child Neurol
.
2015
;
57
(
10
):
899
918
48
Heuser
KM
,
Jaekel
J
,
Wolke
D
.
Origins and predictors of friendships in 6- to 8-year-old children born at neonatal risk
.
J Pediatr
.
2018
;
193
:
93
101.e5
49
Ritchie
K
,
Bora
S
,
Woodward
LJ
.
Peer relationship outcomes of school-age children born very preterm
.
J Pediatr
.
2018
;
201
:
238
244
50
Reyes
LM
,
Jaekel
J
,
Heuser
KM
,
Wolke
D
.
Developmental cascades of social inhibition and friendships in preterm and full-term children
.
Infant Child Dev
.
2019
;
28
(
6
):
e2165
51
Hosozawa
M
,
Cable
N
,
Kelly
Y
,
Sacker
A
.
Gestational age on trajectories of social competence difficulties into adolescence [published online ahead of print March 2, 2021]
.
Arch Dis Child
.
doi:10.1136/archdischild-2020-321317
52
Eryigit-Madzwamuse
S
,
Strauss
V
,
Baumann
N
,
Bartmann
P
,
Wolke
D
.
Personality of adults who were born very preterm
.
Arch Dis Child Fetal Neonatal Ed
.
2015
;
100
(
6
):
F524
F529
53
Baumann
N
,
Tresilian
J
,
Bartmann
P
,
Wolke
D
.
Early motor trajectories predict motor but not cognitive function in preterm- and term-born adults without pre-existing neurological conditions
.
Int J Environ Res Public Health
.
2020
;
17
(
9
):
E3258
54
Husby
IM
,
Skranes
J
,
Olsen
A
,
Brubakk
AM
,
Evensen
KA
.
Motor skills at 23 years of age in young adults born preterm with very low birth weight
.
Early Hum Dev
.
2013
;
89
(
9
):
747
754
55
Poole
KL
,
Schmidt
LA
,
Missiuna
C
,
Saigal
S
,
Boyle
MH
,
Van Lieshout
RJ
.
Motor coordination difficulties in extremely low birth weight survivors across four decades
.
J Dev Behav Pediatr
.
2015
;
36
(
7
):
521
528
56
Pyhälä
R
,
Wolford
E
,
Kautiainen
H
, et al
.
Self-reported mental health problems among adults born preterm: a meta-analysis
.
Pediatrics
.
2017
;
139
(
4
):
e20162690
57
Johns
CB
,
Lacadie
C
,
Vohr
B
,
Ment
LR
,
Scheinost
D
.
Amygdala functional connectivity is associated with social impairments in preterm born young adults
.
Neuroimage Clin
.
2019
;
21
:
101626
58
Rogers
CE
,
Sylvester
CM
,
Mintz
C
, et al
.
Neonatal amygdala functional connectivity at rest in healthy and preterm infants and early internalizing symptoms
.
J Am Acad Child Adolesc Psychiatry
.
2017
;
56
(
2
):
157
166
59
Papini
C
,
White
TP
,
Montagna
A
, et al
.
Altered resting-state functional connectivity in emotion-processing brain regions in adults who were born very preterm
.
Psychol Med
.
2016
;
46
(
14
):
3025
3039
60
Rogers
CE
,
Anderson
PJ
,
Thompson
DK
, et al
.
Regional cerebral development at term relates to school-age social-emotional development in very preterm children
.
J Am Acad Child Adolesc Psychiatry
.
2012
;
51
(
2
):
181
191
61
Bäuml
JG
,
Daamen
M
,
Meng
C
, et al
.
Correspondence between aberrant intrinsic network connectivity and gray-matter volume in the ventral brain of preterm born adults
.
Cereb Cortex
.
2015
;
25
(
11
):
4135
4145
62
Meng
C
,
Bäuml
JG
,
Daamen
M
, et al
.
Extensive and interrelated subcortical white and gray matter alterations in preterm-born adults
.
Brain Struct Funct
.
2016
;
221
(
4
):
2109
2121
63
Bickart
KC
,
Dickerson
BC
,
Barrett
LF
.
The amygdala as a hub in brain networks that support social life
.
Neuropsychologia
.
2014
;
63
:
235
248
64
Bickart
KC
,
Hollenbeck
MC
,
Barrett
LF
,
Dickerson
BC
.
Intrinsic amygdala-cortical functional connectivity predicts social network size in humans
.
J Neurosci
.
2012
;
32
(
42
):
14729
14741
65
Bickart
KC
,
Wright
CI
,
Dautoff
RJ
,
Dickerson
BC
,
Barrett
LF
.
Amygdala volume and social network size in humans
.
Nat Neurosci
.
2011
;
14
(
2
):
163
164
66
Reyes
LM
,
Jaekel
J
,
Bartmann
P
,
Wolke
D
.
Peer relationship trajectories in very preterm and term individuals from childhood to early adulthood [published online ahead of print Marcy 29, 2021
.
J Dev Behav Pediatr
.
doi:10.1097/DBP.0000000000000949
67
Spittle
AJ
,
Orton
J
.
Cerebral palsy and developmental coordination disorder in children born preterm
.
Semin Fetal Neonatal Med
.
2014
;
19
(
2
):
84
89
68
Rogers
M
,
Fay
TB
,
Whitfield
MF
,
Tomlinson
J
,
Grunau
RE
.
Aerobic capacity, strength, flexibility, and activity level in unimpaired extremely low birth weight (≤800 g) survivors at 17 years of age compared with term-born control subjects
.
Pediatrics
.
2005
;
116
(
1
):
e58
e65
69
Brylka
A
,
Wolke
D
,
Ludyga
S
, et al
.
Physical activity, mental health, and well-being in very pre-term and term born adolescents: an individual participant data meta-analysis of two accelerometry studies
.
Int J Environ Res Public Health
.
2021
;
18
(
4
):
1735
70
Tikanmäki
M
,
Kaseva
N
,
Tammelin
T
, et al
.
Leisure time physical activity in young adults born preterm
.
J Pediatr
.
2017
;
189
:
135
142.e2
71
Harrison
H
.
Outcomes in young adulthood for very-low-birth-weight infants
.
N Engl J Med
.
2002
;
347
(
2
):
141
143
,
author reply 141–143
72
Montagna
A
,
Nosarti
C
.
Socio-emotional development following very preterm birth: pathways to psychopathology
.
Front Psychol
.
2016
;
7
:
80
73
Finken
MJJ
,
van der Voorn
B
,
Hollanders
JJ
, et al
.
Programming of the hypothalamus-pituitary-adrenal axis by very preterm birth
.
Ann Nutr Metab
.
2017
;
70
(
3
):
170
174
74
Oswald
LM
,
Zandi
P
,
Nestadt
G
,
Potash
JB
,
Kalaydjian
AE
,
Wand
GS
.
Relationship between cortisol responses to stress and personality
.
Neuropsychopharmacology
.
2006
;
31
(
7
):
1583
1591
75
Jaekel
J
,
Wolke
D
,
Chernova
J
.
Mother and child behaviour in very preterm and term dyads at 6 and 8 years
.
Dev Med Child Neurol
.
2012
;
54
(
8
):
716
723
76
Wolke
D
,
Baumann
N
,
Strauss
V
,
Johnson
S
,
Marlow
N
.
Bullying of preterm children and emotional problems at school age: cross-culturally invariant effects
.
J Pediatr
.
2015
;
166
(
6
):
1417
1422
77
Day
KL
,
Van Lieshout
RJ
,
Vaillancourt
T
,
Saigal
S
,
Boyle
MH
,
Schmidt
LA
.
Long-term effects of peer victimization on social outcomes through the fourth decade of life in individuals born at normal or extremely low birthweight
.
Br J Dev Psychol
.
2017
;
35
(
3
):
334
348
78
Day
KL
,
Van Lieshout
RJ
,
Vaillancourt
T
,
Saigal
S
,
Boyle
MH
,
Schmidt
LA
.
Peer victimization in extremely low birth weight survivors
.
Clin Pediatr (Phila)
.
2015
;
54
(
14
):
1339
1345
79
Wolke
D
,
Copeland
WE
,
Angold
A
,
Costello
EJ
.
Impact of bullying in childhood on adult health, wealth, crime, and social outcomes
.
Psychol Sci
.
2013
;
24
(
10
):
1958
1970
80
Liu
Y
,
Mendonca
M
,
Johnson
S
, et al
.
Testing the neurodevelopmental, trauma and developmental risk factor models of psychosis using a naturalistic experiment
.
Psychol Med
.
2021
;
51
(
3
):
460
469
81
Cheong
JLY
,
Doyle
LW
.
An update on pulmonary and neurodevelopmental outcomes of bronchopulmonary dysplasia
.
Semin Perinatol
.
2018
;
42
(
7
):
478
484
82
Lewis
BA
,
Singer
LT
,
Fulton
S
, et al
.
Speech and language outcomes of children with bronchopulmonary dysplasia
.
J Commun Disord
.
2002
;
35
(
5
):
393
406
83
Caskey
S
,
Gough
A
,
Rowan
S
, et al
.
Structural and functional lung impairment in adult survivors of bronchopulmonary dysplasia
.
Ann Am Thorac Soc
.
2016
;
13
(
8
):
1262
1270
84
Msall
ME
,
Park
JJ
.
The spectrum of behavioral outcomes after extreme prematurity: regulatory, attention, social, and adaptive dimensions
.
Semin Perinatol
.
2008
;
32
(
1
):
42
50

Competing Interests

POTENTIAL CONFLICT OF INTEREST: Dr Marlow declares consultancy fees from RSM Consulting, Takeda, and Novartis in the past 3 years outside this study. The other authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURES: Dr Marlow declares consultancy fees from RSM Consulting, Takeda, and Novartis in the past 3 years outside this study. The other authors indicated they have no conflicts of interest to disclose.