Objectives: In response to parental reports of benefits, we have previously documented improved behaviour, mood, cognition and glycaemia following commencement of CSII in an observational study. This study aimed to re-examine these factors in a randomised controlled trial (RCT).
Methods: RCT of youth with T1D for >1y, aged 7-15y at two Australian tertiary paediatric centres. Participants were randomised to ‘intervention’ (immediate CSII start, n=56) or ‘control’ groups (ongoing use of intermittent insulin injections n=45). Baseline assessments of behaviour and mood (Behavior Assessment Schedule for Children), cognition (standardised tests of intelligence, attention, memory & executive functions) and glycaemia (HbA1c) were conducted pre-randomisation and after 4 months (mo). Primary outcome was difference in parent-reported behaviour at 4mo; differences in self-reported behaviour, mood, cognition and glycaemia were secondary outcomes.
Results: Study groups had similar gender ratio, mean baseline age & HbA1c (11.2 vs 11.0y; 7.9% vs 7.8% in intervention & control groups respectively). Significant differences between groups at 4mo (favouring the intervention) were found in parent-reported behaviour problems (Cohen’s d 0.41; p= .048) and HbA1c (7.4% vs 8.0%; p=.001). Mood (parent report) and some aspects of cognition (perceptual reasoning, attention, cognitive flexibility) improved in both groups over the study, but were not different between groups at 4mo (p>0.05). Self-reported behaviour and mood did not differ between groups.
Conclusions: Parent-reported behaviour problems & HbA1c decreased significantly with CSII vs intermittent insulin injections over 4mo. As externalizing behaviour scores have previously been shown to predict mental health and glycaemic outcomes, this may have long-term benefits. Improvement in mood and some aspects of cognition appear to reflect Hawthorn and practise effects respectively rather than any beneficial impact of pump therapy. This study highlights issues germaine to parental attribution of cause and effect in respect to diabetes therapies and reiterates the value of RCT when evaluating such modalities.