In this study, young adults born preterm VLBW and term SGA reported more problems related to mental health, well-being and social relations than normal birth weight controls. The findings were evident across various questionnaires. IQ had an effect on mental health variables in the VLBW group, while in the SGA group the association with mental health problems was still highly significant after adjusting for IQ. This study highlights the long-term impact on mental health, social relations, QoL and self-esteem of being born at low birth weight, thereby extending former knowledge on this important topic.
The VLBW Group
The VLBW participants had higher mean scores than controls on self-reported Anxious/Depressed and Internalizing Problems scales. These findings are in line with other studies [1–3] and with our previous report of a high prevalence of anxiety disorders in young adults in this group . We also diagnosed more attention deficit hyperactivity disorder in the same study using in-depth psychiatric interview; however, in the present study there were no significant group differences compared with controls on the ASR attention subscale. Hence, ASEBA self-report revealed fewer problems in this group than expected from the in-depth interview. This discrepancy is consistent with our follow-up study in the same VLBW group at 14 years of age . Furthermore, dissimilar ratings on subscales of ASEBA parent-reports and self-reports in VLBW young adult populations have been described [2, 3]; hence non-significant group differences on self-reports should be carefully interpreted.
The higher ASR Total Problems score could not be explained by age at assessment, sex or parental SES, while the same score was reduced when adjusted for full-scale IQ in the VLBW group. Moreover, the higher total scores on autistic traits compared with controls became nonsignificant when participants with cognitive disability were excluded. We found that alterations in brain white matter were associated with cognitive measures and mental health variables among VLBW adolescents at 14 years of age . We therefore speculate that in VLBW individuals, reduced intellectual capacity and comorbid impaired mental health may be largely explained by a deviant brain development persisting into adulthood.
Anxiety, inattention and social problems have been suggested to constitute a “preterm behavioral phenotype” . Autistic traits may affect social skills. Risk of autism is associated with perinatal factors . Increased prevalence of autism spectrum disorders and increased rates of disability pension related to autism spectrum disorders has been reported in adolescence and adult low birth weight populations [14, 15]. In this study the VLBW group had more problems related to attention switching than controls on the AQ questionnaire. Social interaction demands compound skills, and the ability to keep track of simultaneous and complex information in a social setting may be affected by problems in attention switching . Attention problems and cognitive disability may therefore contribute to social problems in the VLBW group, as previously suggested by others [41–43]. Less interaction with friends was reported on the ASR Adaptive score in this group, as has also been previously reported in young adult women born at VLBW . Our findings of more anxiety, attention and subtle social problems may indicate that the “preterm behavioral phenotype” may also apply in young adulthood.
Despite expectations, but consistent with the results of the ASR, mental health domain of QoL was affected by being born at VLBW. This is in line with the negative association between internalizing problems and outcome in health related quality of life found among Dutch young adults born at VLBW or born very preterm . On SPPA-R, the VLBW young adults reported lower self- esteem on athletic performance than controls. Others have demonstrated less physical activity among adolescents and young adults born at very low birth weight [20, 45], and that men born very preterm reported lower physical functioning than full-term controls . Except for lower scores in relation to mental health and athletic performance, other domains of QoL and self-esteem did not differ significantly from controls. One possible explanation may be that the VLBW young adults may have adapted to their situation [23, 24].
The SGA Group
In the SGA group, the problem scores were higher than in the control group for a wide range of mental health problems. The group differences overall were large and consistent with our previously reported finding of higher psychiatric morbidity compared with controls, based on diagnostic assessment . Furthermore, the higher problem scores are in line with studies reporting an increased risk of mental health problems in populations of adults who suffered intrauterine growth restriction (IUGR) [7, 10, 46]. However the large difference in self-reported mental health problems between subjects born SGA and the controls in a cohort with relatively modest growth restriction is unexpected, keeping in mind that studies of mental health outcome in IUGR populations have so far produced ambiguous findings [8, 9]. The higher scores compared with controls on mental health problems could not be explained by age at assessment, parental SES or sex. In contrast to the findings in the VLBW group, they were also not explained by a lower full-scale IQ. Thus, abnormal brain development is less likely to cause mental health problems in young adults born SGA at term. Instead, such problems support the hypothesis of a possible intrauterine programming associated with impaired fetal growth . Intrauterine programming includes probable alterations of the endocrine Hypothalamus- Pituitary- Adrenal (HPA) stress-regulating system [47, 48]. We therefore speculate that the high mental health problem scores reported by young adult SGA participants may be partly explained by a reduced tolerance to stress. This may also play a role in the reported low scores related to mental health domains of QoL and the social acceptance scale of self-esteem. Furthermore, health-related QoL in children and adolescents may decrease over time if mental health problems increase . In a recent study of the same population, we found that mental health problems increased significantly in the SGA group between 14 and 20 years of age . Hence, another possible interpretation of the low QoL measures in the SGA group may be increased mental health problems from adolescence to young adulthood.
Mental health and related concepts
In sum, both groups of low birth weight reported psychological distress, problems with attention switching, affected self-esteem in some domains and reduced quality of life in relation to mental health. Indeed, the sub-scales of SF-36 related to mental health may reflect the higher problem scores reported on ASR. Lower quality of life is associated with mental health problems , and health related quality of life may decrease over time if mental health problems increase . Furthermore, in general, self- esteem in adolescence may predict mental health problems in adulthood  and social relations may be negatively affected by mental health problems. Hence, there is reason to believe that quality of life, self-esteem, social relations and mental health problems are associated concepts, and more research is needed to detect how these problems are interrelated in low birth weight populations.
Clinical and research implications
Our findings suggest that low birth weight may be a risk factor for adverse long term mental health outcome irrespective of whether the cause is preterm birth or intrauterine growth restriction. In this study, the impact of being born VLBW on self-reported mental health problems was reduced when adjusting for IQ. Consequently, psychiatric assessment should include a cognitive examination to facilitate prevention and treatment of mental health problems in this group. Furthermore, as discussed, the young VLBW adults reported less mental health problems on self-reported questionnaires than they did during psychiatric interview. Hence, supplementing self-reports with diagnostic assessment, and also information from multiple informants, may increase the validity of mental health assessment in this group. Young adults born SGA at term reported comprehensive mental health problems and reduced mental health domains of QoL. As this group constitutes a relatively large group of people, there is a potentially high attributable risk.
The findings on mental health in the VLBW group add and extend previous international research on prematurely born children and may probably be generalized to VLBW young adults born from the same geographically based year cohorts in Norway and other developed countries. However, due to significant improvements in perinatal care during the last three decades, caution is needed in generalizing our results to infants born with VLBW in more recent years. Thus, further research including continuous long term follow-up is needed to address the external validity of our findings. Our results in the SGA group represent new findings that need to be replicated in independent samples. Hence; in clinical practice as well as in research, continued attention on the long term mental health outcome and associated risk factors seems indicated.
It will be of interest to see if problems are transitional or if they continue as the subjects grow older. As many psychiatric disorders emerge during adolescence, further research on these low birth weight groups should address adolescent as well as perinatal factors that could possibly trigger adult psychopathology.
Strengths and limitations
The strengths of this study are the longitudinal design into young adulthood and the inclusion of both a preterm and a term born group with low birth weight, by that highlighting some possible long term consequences of both prematurity and intrauterine growth restriction. Furthermore, in this study, we used a variety of measures, providing a broad description of how mental health and well-being were perceived by the young adults themselves.
The limited sample size made separate analyses for men and women difficult, and reduced the power to detect small differences between groups; hence negative findings should be interpreted warily. Although significant differences between the low birth weight groups and controls were found across various measures, the significant p-values in the VLBW group were mostly between 0.01 and 0.05 and should therefore be interpreted with greater caution than the overall highly significant p-values in the SGA group. On the other hand it may be emphasized that all findings were coherent, suggesting more mental problems and reduced well-being in both low birth weight groups, compared with controls. A possible selection bias cannot be excluded in the VLBW group, as the proportion of men was higher among those who declined to participate than in those who participated. Most mental disorders seem to be more frequently occurring in women than in men in the adult population . As there is a greater preponderance of women in our study groups, an amplification of mental health problems and associated factors cannot be excluded. Since SGA is a statistically defined concept, this group comprises subjects born both with and without growth restriction. This misclassification is likely to dilute the real effect of being born with IUGR, and is therefore unlikely to explain the substantial differences between the SGA and the control group. The collected information is based solely on self-reports, but as the participants are adults, this is probably the way they would have represented themselves in most settings at this age. We used a pilot version of the AQ and slight changes made during the subsequent completion of the translation process may reduce the ability to replicate our results.