Mental health problems in pre-school children with specific language impairment: Use of the Strengths and Difficulties Questionnaire

The prevalence of mental health problems (MHP) in children with language disorders ranges from 11 to 55%, due to additional disabilities that have a significant relationship to psychosocial difficulties. Specialists assume that children with a selective disorder [selective language impairment without additional developmental disorder (SLI)], would have less MHP. MHP in SLI have often been studied with the lengthy CBCL-questionnaire. It served as gold standard for the validation of a shorter screening instrument for parents, the Strengths and Difficulties Questionnaire (SDQ). The SDQ has not been validated in SLI at the age 4-8 years. Detection of MHP in SLI at this young age by childneurologists and communicative-specialists is important, as treatment started at an early age might have considerable health benefits. The aims of this study were to assess the prevalence of MHP with, and reliability of, the Strengths and Difficulties Questionnaire (SDQ), in SLI at a young age (4-8 vs. 8-10 years) compared to healthy peers. Methods: We included 151 children with SLI (response 58%) and 326 controls (randomly drawn; response 51%) in 3 special education schools and in 4 out of 17 mainstream schools across the north and middle Netherlands. Parents completed an authorized Dutch translation of the SDQ. Results: 71% were boys. The reliability of the Dutch SDQ was good (4-8 vs. 8-10 years, cronbach a 0.80 and 0.84), compared to controls (0.74 and 0.77 resp.). Sensitivity and specificity were similar to that of the UK version (86 and 96% resp.), but at a lower cut-off (14 and above, vs. UK-cut-off 16). Dutch children with SLI had 32% psychosocial problems, versus controls 11%. Differences in MHP between boys and girls in controls, were absent in SLI, except for more pro-social problems in boys with SLI (p>0.001; ES 0.59). No differences between age bands 4-7 and 8-10 were found in reliability and prevalence. Considering the skewed gender distribution, we assessed differences in SDQ scores by gender and by LI (yes/no), using two-way analysis of variance, but no significance was found. Conclusion: The SDQ is useful for the detection of psychosocial problems in boys and girls with SLI, especially with an adjusted cutoff, and findings at ages 4-7 are similar to those at 8-11 years old. The risk of psychosocial problems in “pure” SLI is three times higher than in healthy peers, and is equal for boys and girls. Diagnostic assessment is required for MHP-diagnoses