Bristol Girls Dance Project: a cluster randomised controlled trial of an after-school dance programme to increase physical activity among 11-12 year old girls

Background: Many children do not meet UK physical activity (PA) guidelines. Girls are less active than boys and the age-related decline in activity is steeper for girls. Dance is the favourite form of PA among UK adolescent girls. Participation in after-school dance could significantly contribute to girls’ PA. Therefore, after-school dance may be effective for increasing PA levels. Objectives: To determine the effectiveness and cost-effectiveness of a dance-based intervention to increase the objectively-assessed mean weekday minutes of moderate-tovigorous physical activity (MVPA) of Year 7 girls one-year after baseline measurement.
Design: Two-arm cluster randomised controlled trial and economic evaluation. Year 7 girls in participant schools received a ‘taster’ session and were invited to participate. Up to 33 girls per school were able to participate. Schools were randomly assigned (equal numbers) to intervention or control arms.
Setting: 18 mainstream secondary schools across greater Bristol.
Participants: Year 7 girls in participating schools who could participate in physical education.
Intervention: Nine intervention schools received an after-school dance intervention (40 x 75 minute sessions) underpinned by Self-Determination Theory – attempting to improve intrinsic motivation for being active - and delivered by external dance instructors. Control schools continued as normal.
Main outcome measures: Main outcome was accelerometer assessed mean minutes of MVPA at T2. Measures assessed at baseline (T0), end of intervention (T1), and T0+52 weeks (T2).
Results: Baseline MVPA levels were high. 508 girls were included in the primary analysis, which found no difference in weekday MVPA between trial arms. There was no effect on secondary accelerometer outcomes. Data were subjected to a per protocol analysis and no effect was found. However, at T1 girls who attended dance had 4.61 more minutes of MVPA and 14.27 minutes of light intensity activity between 3-5pm on the days they attended 3 intervention sessions. The intervention was inexpensive at £73 per participant (£63 when excluding dance instructor travel) but not cost effective due to the intervention ineffectiveness. The EQ-5D-Y data was unresponsive to changes in the sample. The process evaluation reported that girls in attendance enjoyed sessions, exertion levels were low during sessions, attendance was low and declined. Fidelity to the session plan manual was low but theoretical fidelity (to Self-Determination Theory) was good. Qualitative information provides information for improving future interventions.
Conclusions: The intervention was enjoyed by participants. However there was no difference in MVPA levels (which were high at baseline) of girls allocated to receive dance when compared to the control.
Funding: The project was funded by the NIHR Public Health Research (NIHR PHR)
programme (project number 11/3050/01). Intervention costs were paid by the Local Authorities in which the study schools were located. Limitations and Future Work: High baseline MVPA levels indicates that the study appealed to an already active cohort, and therefore may not have targeted those most in need of an intervention. Dance is an enjoyable activity for adolescent girls and could be further trialled as a means to increase PA. Research might consider the impact of dividing the intervention period into smaller blocks.