JAMA Netw Open. 2026 Nov 2;9(11):e2646816. doi: 10.1001/jamanetworkopen.2026.46816.
ABSTRACT
IMPORTANCE: Full-service mobile markets, also known as grocery stores on wheels, may improve diet-related outcomes by bringing low-cost, nutritious food to individuals experiencing low incomes and low food access.
OBJECTIVE: To determine the impact of the full-service Twin Cities Mobile Market on dietary outcomes in low-income communities.
DESIGN, SETTING, PARTICIPANTS: This cluster randomized clinical trial was conducted over 3 waves (January 2022 to October 2025) across 12 public or private subsidized housing sites in the Minneapolis-St. Paul, Minnesota, area. Participants were screened for interest in the mobile market. Follow-up assessments began after 6 months of market service. Analyses were finalized September 2026.
INTERVENTION: Sites were randomized to weekly market service or waiting list control groups.
MAIN OUTCOMES AND MEASURES: Primary outcome: diet quality (Healthy Eating Index-2020 via 24-hour recalls). A priori secondary outcomes: dietary intake (fruit and vegetable, energy, sodium, and percentage of calories from added sugar), food security (18-item US Department of Agriculture Food Security Screening Module), and fruit and vegetable servings procured for the home (validated tracking tool). Adjusted intention-to-treat (ITT) generalized estimating equations were used to assess outcome differences between treatment groups.
RESULTS: Among 262 adults who enrolled in the mobile market study, 242 participants (mean [SD] age, 66.5 [11.5] years; 156 women [64.5%]) across 12 sites completed baseline assessments and were randomized to mobile market service or waiting list control groups. There were 137 individuals (56.6%) who received Supplemental Nutrition Assistance Program (SNAP) benefits. Sites had a mean (SD) of 18.2 (4.4) weekly transactions (mean of 110 of 1071 households [10.3%]). Of 109 intervention site participants, 47 individuals (43.1%) shopped at the mobile market 1 or more times in implementation months 5 to 6, while 90 of 94 individuals who reported ever using the market (95.7%) reported that they would recommend the market service. Participant retention was 218 of 242 participants (90.0%). In ITT models, the primary outcome of diet quality (adjusted mean difference, 1.25 [95% CI, -1.60 to 4.10]) among 215 participants and 6 of 7 secondary outcomes (fruit and vegetable, energy, and added sugar intake; food security and food security improvement; fruit and vegetable procurement) were not significantly different between groups. Intervention site participants consumed an adjusted mean difference of 397 mg (95% CI, -636 to -158 mg) less sodium daily compared with participants in the waiting list control group. Main findings were consistent across 4 sensitivity and supplemental analyses, except the per-protocol sensitivity analysis, which found that participants in the intervention group consumed an adjusted mean difference of 2.35% (95% CI, 0.66%-4.04%) more calories from added sugar compared with the control group.
CONCLUSIONS AND RELEVANCE: In this study, satisfaction with the mobile market was high, but the mobile market did not improve diet quality or most outcomes examined. The mobile market was associated with a meaningful sodium reduction that could improve population-level health.
TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT05672186.
PMID:42975406 | DOI:10.1001/jamanetworkopen.2026.46816
