Exploring the double deficit hypothesis in Spanish schoolchildren using latent profile analysis

Abstract

Background Dyslexia is a neurodevelopmental disorder caused by deficits in phonological awareness (PA). According to the Double Deficit Hypothesis (DDH), Rapid Automatic Naming (RAN) deficit is another potential independent cause. However, the evidence from studies with languages that vary in orthographic consistency is mixed.

Methods The present study aims to investigate the heterogeneity of profiles based on PA and RAN in a sample of Spanish kindergarten children -a consistent orthography- before they start reading instruction (Time 1) to see whether these profiles could predict reading difficulties 2 years later (Grade 2, Time 2). At T1, 795 children were assessed on PA, RAN, letter knowledge (LK), and Verbal Short-term memory (VSTM). At T2, 373 children were reassessed on reading measures. To avoid arbitrary cut-points, profiles were established through Latent Profile Analysis (LPA).

Results Four profiles emerged at both T1 and T2. No single PA or RAN profiles were found. Only the group of low performers corresponded to the Double Deficit profile. The T1 Low-performance group exhibited higher stability than the other groups: 92.5% of the low performers in kindergarten demonstrated poor reading ability in Grade 2, supporting the DDH prediction that children with deficits in both PA and RAN would be more severely impaired than the rest of the profiles. LK showed a relevant contribution in predicting impairment severity.

Conclusions Partial evidence supporting the DDH was obtained. The LPA proved to be appropriate for risk identification under a dimensional conceptualization.