. 2019 Mar 5;22(9):1704–1716. doi: 10.1017/S1368980019000144

Table 2.

Intervention components, measures, statistical models and estimated effects on diet and health outcomes of the studies included in the current review of food pantry-based interventions in the USA

Study	Intervention components	Data collection method	Outcomes	Statistical models	Covariates adjusted	Estimated effectiveness of food pantry-based interventions
Wilson et al. (2017)⁽ ¹⁸ ⁾	Allocation of protein bar (front or back), boxed or unboxed (front–boxed, front–unboxed, back–boxed, back–unboxed) in relation to clients’ choice	Researchers’ observation	Take rate (the proportion of clients that selected the targeted item), binary choice ratios	Logistic regression	Order and packaging	Nudges increased uptake of the targeted food. The findings also hold when authors controlled for a potential confounder. Low-cost and unobtrusive nudges can be effective tools for food pantry organizers to encourage the selection of targeted foods (OR of intervention=4·739; 95% CI 2·269, 9·898; P<0·001)
Caspi et al. (2017)⁽ ²⁰ ⁾	6-week cooking and nutrition education class	Telephone interview or face-to-face (e.g. 24 h dietary recall)	Cooking skills	Paired t test, intent-to-treat analysis		Participants demonstrated improved cooking skills scores post-intervention (P=0·002). This study provides some evidence that improvements in diet and skills can be demonstrated with minimal intervention
Martin et al. (2016)⁽ ²⁹ ⁾	Providing fresh food (including fruits, vegetables, meat and dairy), assistance with housing, education, health care, employment and other basic needs; asking participants to attend the case management meeting	Questionnaire and measurement	Household food security, self-efficacy for food security scale	Hierarchical linear modelling, bivariate analyses, Cronbach α test	Age, gender, Freshplace participation	There was no significant difference in the rate of very low food security in the Freshplace intervention group (51·8%) compared with the controls (47·8%). Self-efficacy was significantly inversely associated with very low food security (P<0·05). Being in the Freshplace intervention (P=0·01) and higher self-efficacy (P=0·04) were independently associated with decreased very low food security
Clarke and Evans (2016)⁽ ³⁰ ⁾	All pantries received additional free vegetables; in addition, pantries in the other conditions received recipes and food-use tips	Telephone interview	Preparing vegetables, methods of cooks, uses of fats and salt			Among the experiment’s ten vegetables, some were used twice as often as others. Cooks practised a narrow repertoire of preparation methods, dominated by boiling and steaming, across most of the vegetables. Fats and salt were often added to boiled and steamed preparations
Seligman et al. (2015)⁽ ²¹ ⁾	Diabetes-appropriate food, blood sugar monitoring, primary care referral, self-management support (recipes and cooking tips)	Blood glucose testing and HbA1c testing	HbA1c, diabetes self-management behaviours, diabetes self-efficacy, adherence, participants’ satisfaction with the food box	t test, χ ² test	Age, sex, race/ethnicity, education, language, site	Improvements were seen in pre–post analyses of glycaemic control (HbA1c decreased from 8·11 to 7·96%; P<0·01), fruit and vegetable intake (which increased from 2·8 to 3·1 servings/d; P<0·01). Among participants with elevated HbA1c (at least 7·5%) at baseline, HbA1c improved from 9·52 to 9·04% (P<0·001). The proportion of participants with very poor glycaemic control (HbA1c>9%) declined from 28 to 25% (P<0·10)
Martin et al. (2013)⁽ ²² ⁾	Client-choice pantry, monthly meetings with a project manager to receive motivational interviewing, targeted referrals to community services	Self-report and questionnaire (e.g. Block Food Frequency Screener)	Food security, self-sufficiency, fruit and vegetable consumption	Multivariate regression model, bivariate analysis, χ ² test, t test, general linear mixed model	Gender, age, household size, income, presence of children in the household	At baseline, half of the sample experienced very low food security. Over 1 year, Freshplace members were less than half as likely to experience very low food security, increased self-sufficiency by 4·1 points, and increased fruits and vegetables by 1 serving/d compared with the control group (P<0·01 for all outcomes)
Flynn et al. (2013)⁽ ²⁴ ⁾	Provision of plant-based recipes that use extra virgin olive oil (6 weeks of cooking classes), consulting dietitians	Questionnaire and measurement	Grocery receipt for all foods purchased, BMI, waist circumference	t test, Wilcoxon signed-rank test		Total variety of vegetables and fruit intake increased (P<0·01 for both). Grocery receipts showed a decrease in purchases of meat, carbonated beverages, desserts, snacks and total groceries (all P<0·01). Food insecurity score decreased from baseline (P<0·01), as did BMI (P=0·05)
Yao et al. (2013)⁽ ³¹ ⁾	Trained volunteers shared whole-grain messages orally, then provided recipe with a bag of ingredients for clients	Telephone survey	Consumption of whole-grain foods, self-efficacy in choosing and preparing foods	Descriptive statistics, Cronbach’s α coefficient, logistic regression	Age, meal preparation experience, family size and number of children	Both perception of whole-grain consumption and self-efficacy improved significantly for the intervention group compared with the control group (P=0·001 and P<0·03, respectively).
Martin et al. (2012)⁽ ²⁵ ⁾	Client-choice pantry, monthly meetings with a project manager to receive motivational interviewing, and targeted referrals to community services, 6-week cooking class, nutrition education	Questionnaire	Food security, self-sufficiency, diet quality	t test, bivariate analysis, χ ² test		Over 3 months, Freshplace members had larger change than the comparison group in food security scores (1·6 v. 0·7 points; P<0·01) and fruit and vegetable intake (1·9 v. –1·4 points; P<0·01)
Clarke et al. (2011)⁽ ²⁶ ⁾	Recipes and food tips, vegetables	Questionnaire and telephone interview	Consumption of fresh vegetables, food-use booklet retention and use	χ ² test, t test, ANOVA	Language preference, restaurant/fast-food places, household size, employment status	Results demonstrated benefits of tailoring over both generic and control conditions and uncovered the degree of tailoring that produced the largest effects (P<0·001). The intervention addressed recipients’ immediate and concrete decisions about healthy eating, instead of distant or abstract goals like prevention of illnesses. The study documented per-client costs of tailored information. Results also suggested that benefits from social capital at sites offering a health outreach may exceed the impact of message tailoring on outcomes of interest
Biel et al. (2009)⁽ ²⁷ ⁾	Display the message about the associations between diet and health, cooking demonstration, provision of fresh food	Staff registration	Number of clinic visits			Pantries and nearby clinics can be brought into collaboration to meet common goals in preventing diet-related illnesses and helping people with such conditions effectively access needed health care. Clinics can effectively partner with food pantries, an overlooked resource for health promotion
Eicher-Miller et al. (2009)⁽ ³² ⁾	Food stamp nutrition education mainly including MyPyramid, food groups, food safety, shopping behaviours and resource management, and wellness	Questionnaire	Food insecurity, food insufficiency	χ ² test, Satterthwaite t test	Food sufficiency pre-test score, employment	Food insecurity and food insufficiency in the experimental group compared with the control group were significantly improved (P=0·03 and P=0·04, respectively)
Greder et al. (2007)⁽ ²⁸ ⁾	Nutrition education programme	Questionnaire	Diabetes, heart disease, high blood pressure, asthma, allergies			A nutrition education programme designed to meet specific nutrition- and health-related needs of pantry participants was developed. Implications include training pantry staff about chronic disease and its relationship to nutrition, identifying pantry foods that provide positive health benefits, and developing consumer publications focused on selecting and preparing pantry foods when one has chronic disease
Miyamoto et al. (2006)⁽ ³³ ⁾	Nutrition education materials focusing on choosing nutritious foods and safe food handling	Oral survey	Food safety, proper hand washing			Of the adult participants, 50% reported preparing the dishes at home; 20% reported enjoying the food, did not prepare the recipe, but intended to do so in the future. Participants reported increasing the frequency of hand washing before preparing and consuming food

HbA1c, glycated Hb.