A gradient relationship between low birth weight and IQ at age 6 years.

To test for a suspected adverse effect of low birth weight (< or = 2500 g) on IQ at age 6 years in two socioeconomically disparate populations. Nonconcurrent prospective study. An urban, predominantly disadvantaged population and a suburban middle-class population in southeastern Michigan. Low-birth-weight (N = 473) and normal-birth-weight (N = 350) subjects, randomly selected from the 1983 through the 1985 newborn lists of two major hospitals, one serving an urban and the other a suburban population. The Wechsler Intelligence Scale for Children-Revised to measure IQ. The mean full-scale IQ of low-birth-weight children was 4.9 points lower than that of normal-birth-weight children, controlling for population site, maternal IQ, maternal education, and race (95% confidence interval [CI], 3.0 to 6.8). There was no evidence of low-birth-weight interaction with population site, and the same IQ difference was detected in both populations. In the urban population, low birth weight was associated with an increased risk for IQ below 85 (odds ratio, 2.2; 95% CI, 1.3 to 3.7). In the suburban population, low birth weight was associated with an increased risk for IQ below 100 (odds ratio, 2.0; 95% CI, 1.2 to 3.4). A gradient relationship with full-scale IQ was observed, with the largest deficit in those born weighing 1500 g or less, an intermediate deficit in those born weighing 1501 through 2000 g, and the least pronounced deficit in those born weighing 2001 through 2500 g. The overall effect of low birth weight represents an increase of approximately 10% of low-birth-weight children with an IQ of more than 1 SD below the population's mean. Children born weighing 2000 g or less bore the major burden of the IQ deficits.