Birth weight and stuttering: Evidence from three birth cohorts.

PURPOSE Previous studies have produced conflicting results with regard to the association between birth weight and developmental stuttering. This study sought to determine whether birth weight was associated with childhood and/or adolescent stuttering in three British birth cohort samples. METHODS Logistic regression analyses were carried out on data from the Millenium Cohort Study (MCS), British Cohort Study (BCS70) and National Child Development Study (NCDS), whose initial cohorts comprised over 56,000 individuals. The outcome variables were parent-reported stuttering in childhood or in adolescence; the predictors, based on prior research, were birth weight, sex, multiple birth status, vocabulary score and mother's level of education. Birth weight was analysed both as a categorical variable (low birth weight, <2500g; normal range; high birth weight, ≥4000g) and as a continuous variable. Separate analyses were carried out to determine the impact of birth weight and the other predictors on stuttering during childhood (age 3, 5 and 7 and MCS, BCS70 and NCDS, respectively) or at age 16, when developmental stuttering is likely to be persistent. RESULTS None of the multivariate analyses revealed an association between birth weight and parent-reported stuttering. Sex was a significant predictor of stuttering in all the analyses, with males 1.6-3.6 times more likely than females to stutter. CONCLUSION Our results suggest that birth weight is not a clinically useful predictor of childhood or persistent stuttering.