Mobile grocery market at low-income housing did not improve overall diet quality, but sodium intake fell — randomised trial

Bringing a full-service grocery store on wheels to low-income housing did not significantly improve residents’ overall diet quality, according to a randomised trial published in JAMA Network Open on 3 November 2026.

The study

• Design: cluster randomised clinical trial of the Twin Cities Mobile Market at 12 subsidised or public housing sites in Minneapolis–St Paul, with sites assigned to receive the market or to a waitlist control.
• Participants: 242 adults, assessed at baseline and after six months of market service using three 24-hour dietary recalls.
• Primary outcome: diet quality on the Healthy Eating Index (HEI-2020).

Results

• The HEI-2020 score did not improve significantly (adjusted mean difference 1.25 points; 95% CI −1.60 to 4.10).
• Most secondary outcomes — fruit and vegetable intake or purchases, energy intake, added sugar and food security — did not change differently between groups. Food security improved in both groups.
• Daily sodium intake was 397 mg lower in the intervention group than in controls (95% CI −636 to −158 mg).

Limitations

Use of the market was modest, which reduced statistical power; data collection was affected by seasonality; and results from one urban area may not apply elsewhere. Dietary intake was self-reported through recalls, and the sodium finding was one of several secondary outcomes. The trial suggests that improving physical access to food alone may not be enough to change overall diet quality.