Feasibility and Response to the San Diego County, California, Supplemental Nutrition Assistance Program (SNAP) Agency Sending Food and Nutrition Text Messages to All Participants: Quasi-Experimental Web-Based Survey Pilot Study

Abstract

Background The Supplemental Nutrition Assistance Program (SNAP) provides over 40 million Americans with money for food without typically providing participants with food or nutrition information. Educational SMS text messages can reach large numbers of people, and studies suggest SNAP participants appreciate nutrition education and have access to mobile phones. Objective Using a pre-post intervention design, we assessed the feasibility of, and program satisfaction and outcomes resulting from, the San Diego County, California SNAP agency sending monthly food and nutrition education SMS text messages to all SNAP participants to increase fruit and vegetable purchasing and consumption. Methods We developed and sent 5 behavioral science–informed SMS text messages with links to a project website in English and Spanish with information about selecting, storing, and preparing seasonal fruits and vegetables. The San Diego County SNAP agency sent monthly texts to ~170,000 SNAP households from October 2020 to February 2021. SNAP participants completed web-based surveys in response to a text invitation from the SNAP agency in September 2020 (baseline, n=12,036) and April 2021 (follow-up, n=4927). Descriptive frequencies were generated, and adjusted multiple linear mixed models were run on a matched data set of participants that completed both baseline and follow-up surveys (n=875) assessing pre- or postattitudes, behaviors, knowledge, and self-efficacy. We used adjusted logistic regression models to assess differences between the matched (n=875) and nonmatched (n=4052) participants related to experiences with the intervention (questions asked only at follow-up). Results After the intervention, matched participants reported significant increase in knowing where to get information about selecting, storing, and preparing fruits and vegetables (3.76 vs 4.02 on a 5-point Likert scale with 5=strongly agree, P<.001); feeling good about participating in SNAP (4.35 vs 4.43, P=.03); and thinking the CalFresh program helps them eat healthy (4.38 vs 4.48, P=.006). No significant pre- or postdifferences were found in fruit or vegetable consumption, though most participants at follow-up (n=1556, 64%) reported their consumption had increased. Among the sample that completed the follow-up survey only (n=4052, not including 875 participants who completed follow-up and baseline), 1583 (65%) and 1556 (64%) reported purchasing and eating more California-grown fruits and vegetables, respectively. Nearly all respondents appreciated the intervention (n=2203, 90%) and wanted it to continue (n=2037, 83%). Conclusions SNAP can feasibly provide food and nutrition messages via text to participants. A monthly text campaign was well received by responding participants and improved some measures of their self-reported knowledge, self-efficacy, produce consumption, and perceptions of SNAP participation. Participants expressed interest in continuing to receive texts. While educational messages will not solve the complex food and nutrition challenges confronting SNAP participants, further work should employ rigorous methods to expand and test this intervention in other SNAP programs before considering to implement it at scale.