Mental health and neural correlates in the preterm brain from adolescence to young adulthood: A longitudinal MRI study
MetadataShow full item record
In the past years, very low birth weight (VLBW) has been associated with deviant brain development, increased risk of experiencing psychiatric disorders and cognitive problems. However, there is still insufficient data explaining the relationship between brain morphology and psychiatric symptoms. In this longitudinal study in preterm-born VLBW individuals, we investigated structural brain changes measured with MRI and its relationship with psychiatric disorders and general cognitive abilities during adolescence and young adulthood. We studied a VLBW cohort and a control group recruited at birth. Structural brain MRI assessments were conducted at three measurement points (15, 19 and 26 years of age) and were analyzed using the FreeSurfer automated software. This software allows the extraction of volumes of cortex and subcortical brain structures and cerebellum as well as cortical thickness and surface area. At 26 years of age, the FreeSurfer software was complemented with TRACULA (TRActs Constrained by UnderLying Anatomy), a software extension for extracting and analyzing white matter (WM) pathways in an automated manner. Psychiatric diagnoses were assessed by senior clinicians at 14 and 19 years using the semi-structured interview Schedule for Affective Disorders and Schizophrenia for School-age Children (KSADS). Diagnoses were established according to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV). The VLBW group was divided in two according to diagnostic change from 14 to 19 years: healthy/becoming healthy and persisting/developing diagnosis. The healthy/becoming healthy VLBW subgroup included those individuals who either were healthy at both measurement points or were diagnosed as healthy at 19 years old. The persisting/developing diagnosis VLBW subgroup included those individuals who either presented a diagnosis or subclinical diagnosis at both measurement points or were diagnosed at 19 years old. At the interview, general psycho-social functioning was rated using the Children’s Global Assessment Scale (CGAS). In order to further assess psychiatric symptoms, the Achenbach System of Empirically Based Assessment (ASEBA) was used, with Youth Self-Report (YSR) at 14 years, and Adult Self Report (ASR) at 19 years. ADHD symptoms were measured with ADHD Rating Scale-IV (ADHD-RS-IV) (mother-report) at both 14 and at 19 years. A senior neuropsychologist assessed cognitive abilities at 19 years, with Wechsler Adult Intelligence Scale, 3rd edition (WAIS-III). Compared with term-born peers, VLBW participants had smaller volumes of cerebellar WM, subcortical gray matter (GM), thalamus and extensive cortical GM volume reductions at 15 and 19 years. At 26 years, they had widespread cortical thickness deviations and surface area reduction in all lobes, and WM abnormalities in forceps minor and uncinate fasciculus. WM abnormalities in the foceps minor mediated cortical abnormalities in the endpoint regions of these pathways, the prefrontal and orbitofrontal regions. At 15 and 19 years of age, VLBW participants who either had or developed psychiatric disorders during adolescence had smaller volumes of cerebellar GM/WM and subcortical GM compared with both controls and healthy/becoming healthy VLBW adolescents, whereas no differences in these areas were found between the last two groups. We did not find any differences between the persisting/developing diagnosis VLBW subgroup and the healthy/becoming healthy VLBW subgroup in cortical GM volumes. Poorer general psychosocial functioning in the entire VLBW group was associated with smaller volumes of cerebellar GM/WM and subcortical GM at 15 and 19 years of age, and with smaller volumes of thalamus and occipital and parietal cortex at 15 years. Inattention symptoms correlated with smaller volumes of cerebellar GM and occipital cortex at 15 years and with occipital and parietal cortex at 19 years. VLBW participants who had or developed psychiatric disorders during adolescence had significantly lower IQ scores at 19 years than VLBW adolescents who were or became healthy during the same period, and, overall, the VLBW group had lower IQ scores than the control group. However, after controlling for the effect of IQ, we still found smaller volumes in parietal cortex, thalamus and cerebellum in the VLBW subgroup with persisting/developing psychiatric diagnosis compared with term-born peers, and in the cerebellum compared with healthy/becoming healthy VLBW adolescents. After correcting for the effect of IQ, poorer general psycho-social functioning in the entire VLBW group was still associated with smaller volumes of occipital and parietal cortex at 15 years and with smaller cerebellar GM volume at 19. At 26 years, higher IQ scores at age 19 years correlated with larger cortical area in frontal, parietal, and occipital regions, and with better WM microstructure integrity in the right uncinate fasciculus. In summary, our results suggest that preterm birth with VLBW has long lasting consequences on brain structure that tend to persist with age. Psychiatric disorders and symptoms were associated with smaller structural brain volumes, especially with regard to subcortical structures and cerebellum. Although lower IQ scores were found especially in VLBW participants who had or developed psychiatric disorders during adolescence, poorer cognitive abilities could not explain the overall higher psychiatric morbidity in preterm-born VLBW adolescents compared with peers born at term.