Evaluation of the feasibility and construct validity of a novel method to measure household fruit and vegetable procurement in low-income community settings.

Background: Many interventions focus on improving fruit and vegetable (FV) procurement for those experiencing low incomes or food insecurity. To address current FV procurement measurement limitations, a novel FV Procurement Tool was developed.

Objective: The main objectives are to describe the tool and evaluate the tool's feasibility and construct validity.

Design: This cross-sectional study used baseline data from the ongoing Mobile Market Cluster Randomized Trial being conducted in the Twin Cities Metropolitan area of Minnesota.

Participants/setting: In 2023, participants (mean=64 years) residing in subsidized housing completed surveys, three 24-hour dietary recalls, and recorded 4 weeks of FV procurement using a novel tool submitted weekly by mail. The tool captured FV items and amounts procured for the home from any type of procurement location, including food pantries. For inclusion in analysis, wave 1 and 2 trial participants (N=161) must have completed at least 3 of 4 FV procurement booklets; 87% (n=140) met this threshold.

Main outcome measures: Main outcomes were Healthy Eating Index-2020 (HEI-2020) score, HEI-2020 Total Fruit and Vegetable component scores, and daily FV servings consumed.

Statistical analyses performed: Descriptive statistics assessed feasibility data (e.g., booklet completion rates) and FV servings by location/source data. FV Procurement Tool's construct validity was assessed with adjusted linear models testing associations between average weekly servings procured of FV, fruit, and vegetables and dietary and survey measures.

Results: On average, 58% of weekly FV servings procured were from a supermarket and 19% from a food pantry. In adjusted models, each 10-serving increase in FV procurement was significantly associated with a higher HEI-2020 score by 1.8 points (95% CI: 0.6,2.9), HEI Total Fruit component score by 0.2 points (95% CI: 0.1,0.4), and daily FV intake by 0.2 servings (95% CI: 0.1,0.3). No other measures were significantly associated with FV servings procured.

Conclusions: This novel FV Procurement Tool may be a feasible method for comprehensively measuring various ways FV enter households. Construct validity was observed for most diet quality and dietary intake measures examined.