A six-week community programme on sugary drinks was linked to a small BMI z-score difference at 12 months

Design

Cluster-randomised trial, 11 community sites, pair-matched

Population

437 children aged 8 to 13 and their carers; 53.7% overweight or obese at baseline

Primary outcome

Change in BMI z score from baseline to 12 months

Effect

Between-group difference -0.19 (95% CI -0.38 to -0.01; P = 0.04)

H2GO! was a cluster-randomised trial across 11 Boys & Girls Club sites in Massachusetts, with sites pair-matched and randomised to the intervention or usual programming. It enrolled 437 children aged 8 to 13 and their carers; at baseline 16.9% had overweight and 36.8% obesity. The intervention was 12 group sessions over 6 weeks, run by trained club staff, aiming to cut sugar-sweetened beverages and raise water intake through sessions, youth-created stories and parent-child activities.

The primary outcome was 12-month change in BMI z score. It fell by 0.11 (95% CI -0.25 to 0.02) in the intervention group and rose by 0.08 (-0.04 to 0.20) in the comparison group; neither change alone was significant, but the between-group difference was -0.19 (95% CI -0.38 to -0.01; P = 0.04). Differences in sugary drink and water intake were not significant.

The interval touches zero, intake did not change measurably, and the setting is a US community club. It suggests community programmes may nudge BMI but is weak evidence for a specific mechanism, and it does not tell a family physician what to do in a consultation.

• Ask about sugary drinks in children with overweight, but do not expect advice alone to shift weight much.
• The between-group BMI z-score difference was -0.19, with a confidence interval just excluding zero.
• The programme was run by community staff, not by doctors, in groups.
• Know whether your area has a community programme for families before you refer.

Why it matters

It adds randomised data on a community-based approach to childhood weight, outside the consulting room.

Don't overread it

The result was borderline, the proposed mechanism (less sugary drink) was not shown, and the setting was a US community club.

The statistics, in plain English

The lower end of the interval (-0.38 to -0.01) sits barely below zero, so the result is borderline. Neither group changed significantly on its own, and the claimed difference comes from the comparison of the two.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.