Outcomes of Early Language Delay

TLDR

Parent‑based assessments of vocabulary, grammar, nonverbal ability, and abstract language use were collected at age 2 from 8,386 twins, with children scoring below the 10th percentile classified as early language delay and later compared to typical peers, while language impairment at ages 3 and 4 was defined as falling below the 15th percentile on at least two of three language measures. At ages 3 and 4, 44.1 % and 40.2 % of the early‑delay group, respectively, met criteria for persistent language difficulties versus 7.2 % and 8.5 % of the typical group, but small effect sizes and poor predictive accuracy from 2‑year data, along with lack of association with initial severity, parental education, or ear‑infection history, limited clinical usefulness.

Abstract

Parent-based assessments of vocabulary, grammar, nonverbal ability, and use of language to refer to past and future (displaced reference) were obtained for 8,386 twin children at 2 years of age. Children with 2 year vocabulary scores below the 10th centile were designated the early language delay (ELD) group, and their outcomes at 3 and 4 years were contrasted with the remainder of the sample, the typical language (TL) group. At 3 and 4 years old, children were designated as language impaired if their scores fell below the 15th centile on at least 2 of the 3 parent-provided language measures: vocabulary, grammar, and use of abstract language. At 3 years, 44.1% of the ELD group (as compared to 7.2% of the TL group) met criteria for persistent language difficulties, decreasing slightly to 40.2% at 4 years (as compared to 8.5% of the TL group), consistent with previous reports of frequent spontaneous resolution of delayed language in preschoolers. Although relations between language and nonverbal abilities at 2 years and outcome at 3 and 4 years within the ELD group were highly statistically significant, effect sizes were small, and classification of outcome on the basis of data on 2-year-olds was far too inaccurate to be clinically useful. Children whose language difficulties persisted were not necessarily those with the most severe initial difficulties. Furthermore, measures of parental education and the child's history of ear infections failed to substantially improve the prediction.