This proposal requests support for a four year follow-up study of the effects of very low birthweight (VLBW less than 1.5 kgm) and brain growth failure on outcomes at 8 years of age. A cohort of 230 VLBW born in 1977-1979 was studied longitudinally at five previous times: at birth, at expected term date, at 8 months, at 2 years and at 3 years of age. Data were obtained on demographic and perinatal variables at birth and on growth and disease variables at all previous points in time. Additionally, IQ measures were taken at 2 and 3 years. The proposed study will extend the longitudinal data set to 8 years of age. Follow-up data will be gathered in 1985-87 on physical growth, neuro-cognitive functioning, and psychosocial adjustment. A sample of 500 age matched children from the same geographic area will be randomly selected to serve as controls. We will test hypotheses that: I) VLBW children will manifest at 8 years significantly more adverse outcomes than randomly selected controls on physical growth, neuro-cognitive and psychosocial functioning; II) brain growth failure will have an independent effect on outcome such that VLBW with subnormal brain growth (less than-2SD) at term who did not catch up by 8 months will have significantly poorer physical growth attainment, neuro-cognitive and psychosocial functioning at 8 years when compared to VLBW who achieved normal brain growth by 8 months; III) these differences will hold when gross neurologic impairment, race and social class are controlled. Furthermore, we will examine the proposition that the long term outcomes of VLBW infants are worse when they are also socially disadvantaged and, moreover, that the combined effects of VLBW and social disadvantage exceed significantly the sum of the individual effects. Data will be collected from children and mothers via structured interviews, physical examinations and standardized tests. These data will be merged with previously documented data on the BLBW cohort. Two types of comparisons will be performed: 1) between VLBW children and controls, and 2) between subsets of VLBW children. Multiple regression analysis will be the major statistical approach used to test hypotheses about differences between samples and subsets of VLBW children.
