Aims Tourette syndrome (TS) is increasingly recognised as a neurodevelopmental disorder associated with a spectrum of behavioural problems. We conducted a case-control study to compare children's self-report measures of behavioural symptoms with their parents' ratings.

Methods Our sample consisted of 28 patients (25 males; mean age 14.1 years; range 12.0–18.1 years) diagnosed with TS according to DSM-IV criteria, and a control group of 61 subjects randomly chosen among sex- and age-matched school peers. All parents completed the Child Behaviour Checklist (CBCL) and the Conners' Parents Rating Scales-R (CPRS-R) for ADHD; all children completed the Youth Self Report (the self assessment version of CBCL). We carried out the following comparisons: TS parents vs controls' parents (CBCL and CPRS-R); TS children vs control children (YSR); TS parents vs TS children (CBCL vs YSR), and controls’ parents vs control children (CBCL vs YSR).

Results Children with TS show higher scores on various subscales of the CBCL compared with control subjects (eg, Total Problems: 70.0±8.6 vs 49.1±8.9, p<0.001; Internalising Scale: 58.1±9.7 vs 51.9±9.4, p=0.005; Externalising Scale: 56.5±9.1 vs 47.2±8.1, p<0.001) and consider themselves more symptomatic than sex- and age-matched school peers on the YSR Problems and Competences Scales (eg, Problem Scales: Tought 53.0±4.6 vs 52.2±6.2, p=0.027; Attention 61.2±10.1 vs 54.0±6.2, p=0.001; Competence Scales: Total 42.9±10.2 vs 49.9±9.3, p=0.002; Social 46.0±7.2 vs 48.8±7.2, p=0.033; 2.3±0.6 vs 2.6±0.4, p=0.002). However, their own reports of behavioural problems are significantly lower than their parents' reports (eg, CBCL/YSR total scores: 53.4±11.2 vs 61.0±8.6, p=0.001).

Conclusions Parents' ratings (CBCL) identify more behavioural symptoms than children's self-report measures (YSR). However, self-report measures such as the YSR are useful in highlighting the areas where children with TS feel more problematic than controls, including social interactions and relationships with peers.