Design: Systematic review and meta-analysis of randomised trials
Population: 46 RCTs in youth and young adults (≤25 years) with type 1 diabetes
Primary outcome: HbA1c
Effect: Pooled g 0.12; family-focused g 0.29; full RCTs g 0.07

An updated meta-analysis in Pediatrics pooled 46 randomised trials of adherence-promoting behavioural interventions in people with type 1 diabetes up to age 25, published between 2009 and 2025.

The overall effect on HbA1c was small (Hedges' g 0.12). Pilot trials reported larger effects (g 0.36) than full-scale trials (g 0.07), a pattern that usually signals small-study inflation. Multicomponent programmes did better (g 0.19) than single-focus ones, and family-focused programmes (g 0.29) did better than those aimed at the young person (g 0.05) or the parent alone (g 0.04).

Despite a decade of pumps and sensors, structured behavioural support still adds only a little to glycaemic control on average. Where resources allow one programme, the evidence leans towards one that brings the family in together.

When referring a young person for diabetes self-management support, favour programmes that involve the whole family.
Expect modest HbA1c effects; set goals around routines and shared responsibility rather than a large HbA1c fall.
Treat results from small pilot programmes with caution; full-scale trials showed much smaller effects.
Revisit family roles in diabetes tasks at adolescent transition points.

Why it matters

Technology has not removed the need for behavioural support, but it has not made that support much more effective either.

The statistics, in plain English

Hedges' g is a standardised effect size: 0.2 is conventionally small, 0.5 moderate. An overall g of 0.12 is a small effect; the subgroup comparisons are observational across trials and weaker than the headline estimate.