Late adolescent and adult outcomes of low birthweight and intrauterine growth failure

The majority of studies on adult outcomes of children born with low birthweight (< 2.5 kg) include both term and preterm children. Prior to the 1960's the mortality of preterm infants was extremely high, and low-birthweight populations included mainly borderline preterm and term survivors, but only one report specifies the gestational age of the children at birth (Paz et al, 1995).

The longest follow-up study is by Martyn et al (1996), who reported on the relationship between fetal growth and cognitive function in middle and late adult life. His population included 1576 singleton men and women, aged 48-74 years, born to married mothers between the years 1920 and 1943 in Herefordshire, Preston and Sheffield, England, on whom birth measurements were available, and who were still living in the area at the time of the follow-up study. The population represented 47% of those who were invited to participate in the study at middle age. Only 74 of the participants weighed less than 5.5 pounds at birth and only 84 were born at less than 38 weeks gestation. The study assessed cognitive function, and its decline with age, by measuring the difference between a vocabulary test, which remains stable with age (the Mill Hill test), and the AH 4 test which measures logical, verbal and numerical reasoning and declines with age. Although cognitive function tended to be higher with increasing birthweight, the results of the study revealed no significant association between body size, or body proportion, at birth and cognitive function, or its decline with age. However, subjects who had a larger biparietal diameter at birth, had significantly higher AH 4 scores. This finding persisted even when adjustment was made for the subject's age and social class and when the subjects born before 38 weeks gestation were excluded from analysis. The authors could not explain this finding since no significant relationship between cognitive function and other head measurements (circumference or occipital frontal diameter), or their relationship to other body measurements was found. They concluded that, "by the time the baby reaches adulthood environmental factors in postnatal life may overshadow any effect of the intrauterine experience".

Stein et al (1972) studied the effects of prenatal exposure to famine in Holland during World War II (1944-1945). The study population included 125,000 males born in 7 famine-stricken areas and 11 areas not exposed to famine, on whom psychological and educational tests were performed at induction into the military at 18-19 years of age. Ninety-six percent of the births were located for the study. The authors noted a decrease in mean birthweight and birthweight below 2000 g during the famine, however there were no differences in intelligence, measured by the Raven Progressive Matrices Test, or in the rates of mild or severe mental retardation, between subjects from the famine and control areas. Stein et al concluded that starvation during pregnancy had no effects on intelligence and that there was no clear association between mean birthweight and intelligence. During the famine, births decreased more among the lower than among the higher social classes, whereas after the famine, there was a compensatory increase in births among the lower social classes. Social class effects might thus have affected the mean intelligence scores both during and after the famine. Stein also noted that the population might represent a selective survival of the fittest or that postnatal experiences might have had a compensatory effect on the outcomes.

Douglas and Gear (1976) followed 80 of 163 singleton survivors with birthweight less than 2000 g who participated in the 1959 longitudinal British Birth Cohort Child Development Study. Sixty-seven subjects (84%) were tested at 18 years of age and compared to matched controls with normal birthweight. Although significant differences in academic performance had been noted at 8 years of age, no significant differences in the rates of mental or behavioral handicaps were noted later. At the age of 15 years the low birthweight children had a mean IQ of 93 compared to 97 for the normal birthweight controls, but this difference was not statistically significant. Douglas noted that the results might have been confounded by the fact that, although the groups had initially been matched by social class, the home circumstances of the control families improved over the years, "possibly due to a greater drive and social responsibility among these families". A similar divergence of social circumstances between the low birthweight and control families over time was noted by Illsley and Mitchell (1984).

Nilsen et al (1984) in Norway examined the outcomes of a hospital population of children born 1962-1963 with birthweights < 2500 g, when they were conscripted to the army at the age of 18 years. Twenty-nine children had birthweights below the 10th centile for gestation; two of them were considered unfit for military service. No differences in intelligence scores were noted between the remaining 27 IUGR subjects and controls.

Paz et al (1995) reported on the outcomes of 17-year-old IUGR (< 3rd percentile for gestation) term subjects born in Jerusalem, Israel, who were tested prior to conscription to the army. The IUGR subjects had significantly lower IQ scores when compared to controls, but when the scores were adjusted for perinatal risk factors and socio-demographic status, the differences remained significant for females only. The conclusions that can be drawn from this study are limited by the: small number of IUGR children (30 males and 34 females), the exclusion of severely handicapped children from the army evaluation, and the greater likelihood of females with low education achievement to be exempt from conscription to the army. Significantly more of the IUGR males (40% versus 23% for normal birth weight controls) had low educational achievement defined as education less than 12 years or attending a special educational school. For females, the rates were 15% versus 6%, respectively. IUGR birth had no effect on having an IQ < 85 when multivariate analysis was performed to control for confounding variables such as ethnic origin, parental education, social class and birth order.