Evaluating sleep quality using the CSHQ-Autism

Sleep problems are common in autistic children and adversely impact daytime functioning. The Children's Sleep Habits Questionnaire (CSHQ) [39] was developed from a community-based sample of children and has validated a cut-off score of 41. Katz et al. [50] developed an abbreviated 23-item four-factor version of the CSHQ, which may be useful when assessing sleep in autistic children. However, a cut-off value has not yet been developed.

Our objective was to develop and validate a cut-off for the CSHQ-autism total score in order to identify sleep problems among autistic children. We hypothesized that the derived cut-off value for the CSHQ-autism would perform better than the original CSHQ cut at 41 on validation in a sample of autistic children.

Age-specific cut-off values were developed and validated using receiver operating characteristic analysis.

The derived cut-off values for the CSHQ-autism total score were 34, 35, 33, and 35 for the 2-3, 4-10, 11-17, and 2-17 years age groups, respectively. On validation, all cut-off values performed with moderate to high sensitivity (76.6-82.4%) and moderate specificity (69.1-75.5%), while the original CSHQ cut at 41 had high sensitivity (89.9-93.0%) but low specificity (42.6-57.7%). Using McNemar's tests, the CSHQ-autism had significantly higher specificity but lower sensitivity than the original CSHQ cut at 41 in all age groups.

The CSHQ-autism cut-off values performed better overall than the original CSHQ cut at 41 in a sample of autistic children. The CSHQ-autism cut-off can help identify sleep problems among autistic children.