ABBREVIATIONSCHAID v 2 automatic interaction detection LCA Latent class analysis MPT Moderately preterm AIM The aim of this study was to identify subgroups of children born moderately preterm (MPT) and term with distinctive levels and patterns of functioning, and the perinatal and demographic factors that predict subgroup membership.METHOD A total of 378 children aged 7 years, 248 MPT (138 males, 110 females; gestational age 32-36wks) and a comparison group of 130 children born at term (58 males, 72 females; gestational age 38-41wks), were selected from a community-based cohort study. Latent class analyses were performed on measures of intelligence, verbal memory, attention, executive functioning, and visuomotor and motor skills. v 2 automatic interaction detection analyses were performed to detect associations between the subgroups and predictors.RESULTS Four subgroups differing in levels of performance were identified, with parental education being the only statistically significant determinant of subgroup assignment (p<0.01). The subgroup that performed the most poorly showed an irregular pattern of performance, with specific weakness in attentional skill and relative strength in intelligence and verbal memory. Parental education predicted classification probability in the preterm group (p=0.04) but not in the term group (p=0.15).INTERPRETATION Our results show that the poorer performance of children born MPT reflects a higher proportion of children with below average performance rather than a subgroup with extremely poor performance. They indicate that MPT birth affects neurodevelopmental functioning at early school age only slightly, with effects being largest in such children with low parental education.Moderately preterm (MPT; gestational age [GA] 32-36wks) birth accounts for approximately 7% of all births in Europe and 10% in the USA. 1 Follow-up studies have consistently found that children born MPT, at preschool and school ages, show poorer intelligence, visuospatial reasoning, attention, executive functioning, and motor skills than children born at term. 2,3 Although significant, the differences between the two birth status groups are small. 2,3 For example, van Baar et al. 3 found that the IQ scores of MPT children were within the normal range but on average three IQ points lower than those of term children. As previous studies have gone no further than comparing average performance across a range of functions in the two groups, the basis for these small differences is not yet known. They may reflect extremely poor performances, i.e. in the clinical range, in a small subgroup of MPT children or slightly poorer performances in a higher proportion of MPT children. Furthermore, comparisons of averages provide little insight into the individual differences in performance underlying the group differences. These may reflect different patterns of strengths and weaknesses in different functions or small differences across a range of functions. Understanding both the level and pattern of cognitive and mo...