Main

The survival of very preterm infants has improved during the past two decades and led to an increased interest in their long-term neurodevelopmental outcome. Reported series of these infants have shown major disability rates of 10% to 15%, caused mostly by cerebral palsy and hearing and visual disabilities (1). The majority of the remainder attend normal schools, but one third to one half of these experience behavioral and specific learning difficulties often associated with minor motor impairment (MMI) (2). In the absence of other neurodisability, this syndrome is often referred to as developmental coordination disorder (3). Whereas imaging techniques have clearly shown lesions associated with major neurodevelopmental disabilities, this has not been the case with MMI. There seems to be little if any correlation between cranial ultrasound- or magnetic resonance imaging (MRI)-detected lesions and subsequent functioning at school age despite the high prevalence of such lesions in children born preterm when compared with term control subjects (4, 5). Preterm infants have been shown to have smaller brains, less cortical complexity than term children, and often delayed myelination. Quantitative MRI in adolescents who were born preterm has shown significant associations between caudate and hippocampal volumes and intelligence (6) and hippocampal volumes and everyday memory (7)

MRI is highly sensitive to the age-specific changes in myelination; qualitative assessment of the extent of myelination on MR images is a routine part of assessment of brain maturation (8). Visual assessment of gray-white differentiation, 1H MR spectroscopy and diffusion tensor MRI, and reduction in cortical folding all have been used to show delayed cerebral maturation in preterm infants at term (9, 10). A simpler quantitative technique, T2 relaxometry, provides an objective measurement of tissue characteristics; the T2 relaxation time is a physical property of the selected volume of tissue, which mainly reflects the free water content of the tissue. T2 relaxometry of the hippocampal formations has been used to assess patients with complex partial seizures and has been shown to be more sensitive than visual assessment of MR images (11). T2 relaxometry of the cerebral white matter (CWM) has been described in adult patients with myotonic dystrophy (12) and in children affected by phenylketonuria (13, 14), the 18q syndrome (15), and sexual abuse and drug use (16). Prolonged white matter T2 relaxation times in these patients were considered to represent impaired myelination or gliosis, resulting in an increase in free water concentration. T2 relaxation times have been shown to be high after birth, declining rapidly during the first year of life and achieving adult levels at between 2 and 3 y (17). We hypothesized that MMI in children who are born very preterm is related to evidence of a generalized CWM abnormality rather than only a focal cerebral injury and that similar changes in the hippocampus would reflect IQ deficits.

METHODS

Participants.

All infants who were born before 32 completed weeks in 1991–1992 in the eight hospitals within the Liverpool postal districts were ascertained. Those who died before discharge from hospital or whose mothers were not resident within a Liverpool postal district at the time of birth were excluded.

Initial contact was made with the family doctor via the child's hospital pediatrician to ascertain current health status and school placement. The parents of those children who were alive and attending mainstream schools were then approached to seek consent to take part in the study. When informed consent was obtained, the individual children's schools were then contacted to arrange assessment visits. Most children were tested at their schools, although a few were tested at the Institute of Child Health, Royal Liverpool Children's Hospital, at their parents' request. The study protocol was approved by the local Research Ethics Committee.

Test instruments.

Fine and gross motor skills were assessed using age band 2 of the Movement Assessment Battery for Children (Movement ABC) (18). The test comprises eight items, divided into four subsections: manual dexterity, ball skills, static balance, and dynamic balance. The scoring system for each item is from 0 to 5, ranging from no impairment to severe impairment. The scores for each item are added and converted to percentiles. A score between the fifth and 15th percentiles for age is considered “borderline” impairment, and at or below the 5th percentile “definitely impaired.” In this study, a score at or below the 5th percentile was taken to indicate definite MMI.

General intelligence was measured using the Wechsler Intelligence Scale for Children UK (19). Total, verbal, and performance scores were calculated.

Imaging techniques.

All parents were approached to ask for agreement for their child to undergo cranial MR scanning at age 7 y. When this was obtained from both parent and child, scans were performed on a 0.5 Tesla Philips NT-5 scanner under the supervision of a pediatric imaging team. Scans were assessed qualitatively for signs of perinatal brain injury, including periventricular leukomalacia (PVL), porencephaly, ventricular dilation at any site, and thinning of the posterior portion of the corpus callosum. PVL was recognized by a characteristic triad of abnormalities: abnormally high signal within the periventricular white matter on T2-weighted turbo spin echo (TSE) and fluid attenuated inversion recovery images, loss of periventricular white matter, and focal ventricular enlargement adjacent to regions of abnormal signal intensity.

The scans included a T2 mapping sequence, which allowed measurement of T2 relaxation times from regions of interest in the CWM and hippocampal formations in both cerebral hemispheres. The total scanning time for each patient was just over 20 min, including 4 min and 25 s for the T2 mapping sequence. This technique involved the selection of a single coronal slice, within which the body of the hippocampus and the white matter of the centrum semiovale were represented. By an automated process, 16 separate spin-echo images of this slice were obtained at different echo times, producing a T2 map of the selected slice in which the brightness of each individual pixel represented the calculated T2 relaxation time. Little anatomic information is visible on the T2 map, so regions of interest were drawn manually on the equivalent T1-weighted image, on which there is clear contrast between gray and white matter. For each examination, a total of four regions of interest were drawn: one in the white matter of the right cerebral hemisphere, one in the white matter of the left cerebral hemisphere, one in the right hippocampus, and one in the left hippocampus. These regions of interest were then pasted onto the T2 map, and the mean T2 relaxation time for each region was calculated (Figs. 1 and 2).

Figure 1
figure 1

(A) Region of interest drawn within the cerebral white matter on coronal T1-weighted MRI scan. (B) Region of interest pasted onto T2 map.

Full size image
Figure 2
figure 2

(A) Region of interest drawn in the body of the hippocampus on coronal T1-weighted MRI scan. (B) Region of interest pasted onto the T2 map.

Full size image

This method is an adaptation of the technique that is routinely used for measurement of T2 relaxation times of the hippocampal formations in the investigation of patients with temporal lobe epilepsy. T2 relaxometry is more sensitive than visual evaluation of T2-weighted images in detecting subtle hippocampal sclerosis (9).

Measurements were obtained independently by two radiologists (L.J.A. and G.K.) who were not aware of the motor and psychometric test results. Intraobserver and interobserver reliability of T2 relaxation time measurement were assessed by the Bland-Altman method (20). For intraobserver differences, 95% confidence limits of agreement were −2.4 to 2.4 ms for observer 1, and −1.9 to 1.7 ms for observer 2. For interobserver differences, 95% confidence limits of agreement were −2.4 to 2.8 ms. Statistical analysis of data was carried out using SPSS-10, using parametric or nonparametric tests depending on the distribution of the data.

RESULTS

Of a potential cohort of 382 preterm children identified, 33 had moved out of the area or could not be traced, 18 had died, 29 were attending special schools, 16 parents refused permission for their children to be tested, and six failed to attend assessments after giving consent. The 280 children tested comprised 151 (53.9%) boys and 129 (46.1%) girls. Their mean gestational age was 29.8 wk (range 23–32), and their mean birth weight was 1467 g (range 512–2860); 21.4% were of ≤28 wk gestation, and 3.6% were ≤24 wk; 50.4% were <1500 g birth weight, and 14.6% were ≤1000 g. There were 215 singleton births, 56 twins, and nine triplets. The mean age at testing was 89.8 mo (range 82–101). The total median score on the Movement ABC was 8.5 (IQR 3.1–15.3). Using the fifth percentile to define a MMI group, 86 (30.7%) fulfilled this criterion. No significant differences in the frequency of MMI were seen between the sexes. A significant negative correlation between gestation and birth weight and score on the Movement ABC was seen (Spearman ρ, −0.190, p = 0.001 for gestation; −0.176, p = 0.003 for birth weight). However, a gestational age of ≤28 wk (p = 0.24) or birth weight <1500 g (p = 0.056) was not associated with MMI. Mean (SD) total, verbal, and performance IQs were 89.4 (14.2), 92.9 (13.9), and 87.8 (15.6). Total IQ was correlated significantly with gestational age and birth weight (Spearman ρ 0.264, p < 0.001; 0.229, p < 0.001). The impairment score on the ABC was significantly correlated with total, verbal, and performance IQ (Spearman ρ −0.437, p < 0.001; −0.265, p = 0.02; −0.442, p < 0.001).

A total of 103 of the 280 preterm (PT) children tested at school agreed to have an MR scan. Tables 1 and 2 show descriptive perinatal variables and variables at 7 y for preterm children who were scanned and for those who were not scanned. Table 3 shows the median (IQR) test scores for the ABC and the mean (SD) test scores for the full, performance, and verbal IQs in infants with and without cerebral lesions visible on MRI scans. On qualitative assessment of the MRI scans, 20 children had abnormal scans; 14 had evidence of periventricular leukomalacia (PVL), 14 had ventricular dilation, eight had thinning of the corpus callosum, and two had porencephaly (Table 4). The presence of all lesions except for porencephaly was associated with lower IQs. The ABC scores were significantly impaired only in the presence of marked thinning of the posterior part of the corpus callosum.

Table 1 Characteristics of infants in perinatal period who later did or did not have an MRI scan
Full size table
Table 2 Characteristics of infants with and without MRI scan at 7 y
Full size table
Table 3 Lesions seen in 20 preterm infants on MRI scans
Full size table
Table 4 Test scores in presence or absence of cerebral lesions on MRI scans
Full size table

Table 5 shows T2 relaxation times for the hippocampal formations and CWM for each side for all children scanned. The T2 relaxation times are significantly longer in the CWM in infants with impaired scores on the ABC (<5%). When only infants without visible lesions on scan are considered (Table 5), the differences persist, although at borderline statistical significance. Similar comparisons of IQ scores to T2 relaxation times (Table 6) showed no statistically significant differences for any region, whether total, performance, or verbal IQ was considered or when test subsets were used.

Table 5 T2 relaxation times for right and left hippocampus and right and left CWM in infants with MMI (ABC score <5th centile). a) All cases scanned (n = 103); b) Only cases without visible cerebral lesions on MRI scan (n = 83).
Full size table
Table 6 T2 relaxation times for right and left hippocampus and right and left CWM in infants with IQ above and below 85 (−1 SD).
Full size table

DISCUSSION

Learning and behavioral problems have been widely recognized under a variety of names as common sequelae of preterm birth. MMI is often associated with visuospatial problems, poor mathematical ability, and visual problems such as low contrast sensitivity, which in the absence of major cerebral lesions on imaging suggest a cortical developmental problem. Preterm infants have been shown to have smaller brains and impaired white and gray matter development at term compared with full-term control subjects. In the neonatal period, maturational changes on MR mirror infant behavioral scores (21), and reduction in regional cortical volumes in preterm children at 8 years was associated with lower IQ scores (22).

The diagnosis of MMI is problematic in that several different tests are regularly used and the cut-off scores used to define the condition vary. The definition used in this study has been fairly widely used. Other features such as attention deficit, low IQ, and poor social skills are often seen in conjunction with MMI, although the presence of a general medical condition such as cerebral palsy excludes the diagnosis. It has even been suggested that survival after extreme prematurity is such a general medical condition (23). Although the minor impairments seen in such children are diverse, their overall frequency is similar in reported series (1, 4, 24–28).

This study has confirmed the high prevalence of MMI and lower IQs seen in preterm children without major neurodevelopmental sequelae at primary school age. It has also shown a clear correlation between MMI and IQ and visible lesions on the MRI scans such as PVL, ventricular dilation, and thinning of the posterior part of the corpus callosum. These changes are likely to represent gross loss, atrophy, or gliosis of the periventricular white matter after the perinatal period. Individually, these visible lesions were not related to IQ deficits, even though the score on the ABC and IQ were significantly correlated. This suggests that the lesions that cause IQ deficits and MMI are not always the same. IQ relates to brain size and cortical gray matter volume, which may be impaired independent of previous white matter injury such as PVL (29). Although visible white matter damage in the form of PVL, ventriculomegaly, and thinning of the corpus callosum may produce later reduction in cortical gray matter through disruption of cortical association fibers, it is possible that the lesion represented by prolongation of T2 relaxation time is different and does not affect later brain development. Indeed, it may itself be related to later factors such as poor overall growth. Our results also show that CWM T2 relaxation time is significantly increased in preterm children with MMI. A T2 relaxation time of >89.5 ms at this age predicted MMI with a sensitivity of 80% and a specificity of 60%, compared with any visible MRI lesion with a sensitivity of 32% and specificity of 84%. This difference persisted when cases with visible MRI lesions were excluded, suggesting that a diffuse abnormality of CWM may be responsible for MMI in most of these children. This could represent a subtle disturbance of white matter myelination, which is independent of the macroscopic changes of PVL.

T2 relaxation time reflects the relative proportion of water to other matter such as myelin, and prolonged times could also result from acute events such as cerebral edema after perinatal hypoxia or ischemia as well as demyelination or delay in myelination. These children were 7 y old, and the immediate effects of perinatal injury were unlikely to have persisted. Gliosis and sclerosis may also present with altered T2 relaxation times, although differences in T2 relaxation times were seen when children with visible evidence of old injury on scans was absent. Further elucidation of the exact nature of these changes may depend on the development of faster methods for MR spectroscopy that could be used practically in pediatric research subjects without sedation.

In the first year of life, the composition of CWM undergoes a rapid sequence of change. Progressive myelination results in an increase in lipid and protein content and a reduction in water content. The water content of the brain decreases from 88% at birth to 82% at 6 mo. Myelin eventually constitutes 50% of the weight of white matter in the mature brain (30).

A generalized defect in postnatal myelination could represent a common cause. Such defects are possibly due to injury to axons and oligodendrocyte progenitors by cytokines and free-radicals after perinatal infections or ischemia (31), effects of drugs such as corticosteroids (29), or nutritional deficiencies (32). Identification and modification of these factors may offer the possibility of specific therapies to reduce the burden of disability in children who are born preterm.