TABLE 3.

Baseline characteristic of included studies (n = 12).

	No.	First Author, year	Country	Study design	Target population (n, age, ethnicity)	Setting (level)	Intervention description (components, strategies, and study groups)	Duration	Outcomes measure (scale/methods)	Key findings
Agricultural interventions/Developing countries
Non-randomized controlled trials	1.	Boedecker et al. (51)	Western Kenya	Cross-sectional study (two groups comparisons)	444, Children 12–23 months, women 15–49 years	Community (household)	● Workshops encourage and support communities in planning agricultural activities to improve nutrition, raising awareness on nutrition and healthy diets, identifying poultry raising, and kitchen gardening (especially traditional legumes and leafy vegetables) to support diet diversification. ● The workshops were led by a nutritionist of the country’s Ministry of Health (MoH) to share nutrition contents, Bioversity International and CHVs [I: 5 sublocation (n = 296) and C: 5 sublocation (n = 148)]	12 months	Mean dietary diversity score/DDS, the percent of women and children reaching minimum dietary diversity/MDD, and micronutrient adequacy using mean adequacy ratio/MAR	The intervention significantly effect on children’s mean DDS (the size of treatment effect = 0.7, p < 0.001) and children reaching MDD (the size of treatment effect = 0.2, p < 0.001)
	2.	Carney et al. (43)	Oregon	Intervention study (Pre-post survey)	131, 4–44 years	Community (household)	● The strategies provided by the community meeting included the project materials, such as seeds, and gardening strategies, such as preparing the land, choosing plants, compost, organic approaches for pest control, maintaining the garden, and harvesting the vegetables	43 months	Food security (accessibility), vegetable intake, food safety	In adults, vegetable intake increased from 18.2 to 84.8% (p < 0.001). Children’s vegetable consumption increased from 24.0 to 64.0% (p = 0.003).
	3.	Kansanga et al. (50)	Malawi	Cross-sectional study (two groups comparisons)	914, ≥30 years	Community (household, farmers)	● Using agroecological practices improves farmers’ knowledge, agricultural productivity, and household nutrition. In order to improve agricultural productivity and household nutrition, the project trained smallholder farmers on applying agroecological practices using local resources and a farmer-to-farmer knowledge-sharing approach. (I: agroecology-practicing farming households (n = 514). C: non-agroecology households (n = 400)	60 months	household production diversity and dietary diversity	A significant positive effect was shown in the mean dietary diversity (β = 0.175, p < 0.01) and production diversity of households (β = 0.289, p < 0.01).
	4.	Kazige et al. (48)	Mushing area, Walungu territory, South-Kivu province, eastern DRC (Congo)	Intervention study (Pre-post survey)	Rural household, NR	Community (household)	● In this study, to improve the food security of rural households in South Kivu, residues of four staple crops, including banana, cassava, maize, and common bean, were valorized in two stages: (1) monitoring of the fields of farmers throughout the cropping season to record the weight of crop residues and yields, and (2) evaluation of the potential of staple crop residues for mushroom production. A total of 70 fields were selected for this purpose	10 months	Effects of substrate ingredients (common bean, banana, cassava, and maize) and the additive used to increase production	43.5% of the households consumed common beans as vegetables, and 20.7% sold them at the local market. Maize gave the highest mushroom yields (2.4 kg kg–1)
	5.	Madsen (52)	Malawi	Longitudinal pre-post and delayed intervention	537, 43 years,	Community (farmers)	● Training on soil management, ● Soil compost/manure application, mulching, crop rotation, and agroforestry. Acquaintance of the participants about recipes with new products and the re-introduction of grains and legume crops [I: received intervention from the beginning (n = 428) and C: delayed intervention (n = 109)]	36 months	Food security ● Household Food Insecurity Access Scale-HFIAS (HFIAS)	Crop yield diversity increased in intervention farms from a mean of 2.06 crops to 4.23 crops. The percentage of food security increased from 10% at the baseline to 30% at the end line.
	6.	Nyantakyi-Frimpong et al. (49)	Malawi	Cross-sectional study (two groups’ comparisons)	1000, ≥18 years	Community (household)	Sustainable farming techniques, such as organic composts; intercropping; mulching of stubble; and no-till farming. [I: agroecology-adoption (n = 571), C: non-agroecology-adaption (n = 429)]	NR	Health and well-being (self-reported), food insecurity access (Household Food Insecurity Access Scale-HFIAS)	Agroecological-conscious households were more likely to report ideal health status (OR = 1.37, p = 0.05), According to the average treatment effect, adopters had a 12% higher chance of being in optimal health. Moderate food insecurity (OR = 0.59, p = 0.05) and severe food insecurity (OR = 0.89, p = 0.10) were linked to a lower chance of reporting optimum health status.
Nutritional interventions/Developed countries (poverty regions)
	7.	Ball et al. (46)	Catawba County, North Carolina (United States)	Intervention study (Pre-post survey)	415, NR	Community (farmers)	● Given Farmers’ Market Nutrition Program (FMNP) coupons weekly for 24 weeks to observe the marketing behaviors (I: 415 farmers).	72 months	fresh fruits and vegetables (F and V), quality of fresh F and V, and the ability to purchase food grown locally and increase the FMNP coupon redemption rate	Between 2007 and 2013, the average redemption rate was 51.4% (rank of 10). In 2013, the redemption index increased from 51.3 to 62.9% (rank of 3).
	8.	Breckwich et al. (44)	San Francisco, California (United States)	Intervention study (Pre-post survey)	Customers, NR	Community (stores)	● Advertising the products in the pilot store through promotional giveaways of canvas shopping bags, taste testing, fresh produce, and conducting in-store press events. ● Working with local graphic artists to develop and implement an award-winning media campaign to raise awareness about the program. ● Coverage of the pilot store intervention by several local media sources	6 months	To execute an inventive and sustainable public policy solution to expand community access to healthy food	In 3 years, this partnership achieved a remarkable number of policy-related successes, including youth-led research on the issue of food insecurity, a fruitful pilot Good Neighbor store intervention, community outreach, and education to influence community knowledge and behavior change, the development of a city- and foundation-sponsored initiative to expand the Good Neighbor Program, and state legislation to support similar initiatives across California.
	9.	Dailey et al. (45)	Adams County, Colorado, United States	Intervention study (Pre-post survey)	47 Latino or Hispanic (81.6% Hispanic), NR	Community (household)	● Given vouchers to families for increasing fruit and vegetable consumption	4 months	Food security (accessibility), fruit and vegetable intake, increased ability to purchase healthy, fresh foods	34.2% of participants reported consuming 2–3 servings of fruits and vegetables daily, and two-thirds of the respondents reported four fruits and vegetables per day. Over 40% of respondents reported greater stress related to having enough money to purchase nutritious meals.
	10.	Rollins et al. (47)	Atlanta, Georgia (United States)	Intervention study (Pre-post survey)	11 corner 100 African American corner store customers (≥18 years) with little access to nutritious foods	Community (corner stores)	● Stores sold fresh, healthy products and promoted cleanliness of store labeling the products. A customer intercept survey was administered at five healthy corner store locations to African American customers.	12 months	Assessing the Healthy shopping behavior	80% of customers said that they would purchase healthy food, e.g., fresh and inexpensive vegetable and fruits if sold in corner stores
Randomized controlled trials	11.	Jernigan et al. (53, 54)	Oklahoma, United States	cluster randomized controlled trial	1204, ≥18 years Chickasaw Nation (403 control, 410 intervention) and Choctaw Nation (409 control, 415 intervention)	Community (stores)	● Improve convenience stores and healthy retail strategies recommended by the Institute of Medicine and Centers for Disease Control, including (1) increasing availability, variety, and convenience of healthy foods; (2) placement of point of purchase details; (3) promoting, advertising, and selling nutritious foods; and (4) reducing healthy foods prices measured by NEMS tools to assess objective changes in the nutrition environment of the stores before and after the intervention. (I: two stores received the intervention and C: two stores did not receive the intervention)	9–12 months	Individual-level outcomes: ● Increases in fruit and vegetable consumption and secondary outcomes included consumption of other foods, changes in the perceived food environment, recall of promotions, and reported purchase of healthy foods. Store-level outcomes: ● Availability of healthy foods with an emphasis on ready-to-eat fruits and vegetables, in addition to pricing, placement, and quality measured by NEMS tools	Individual-level outcomes: After the intervention, both control and intervention participants’ daily fruit and vegetable intake stayed low. Following the intervention, both intervention and control participants believed that stores were healthier. Purchases of fruits, vegetables, and other healthy items were linked to higher shopping frequency. Store-level outcomes: There were no variations in the median NEMS scores between intervention and control stores.
Nutritional interventions/Developing countries
	12.	Kang et al. (41)	Ethiopia	cluster randomized controlled trial	1790, 6–12 months	Community (mothers)	● Education of mothers in 12-day nutrition classes focused on child feeding that follows the ’learning by doing approach. I: (n = 876), C: (n = 914)	15 months	Weight-for-age (WAZ), weight-for-length/height z score (WLZ/WHZ), and Length/height-for-age z score (LAZ)	Children in the intervention area had faster growth in length [difference (diff): 0.059 cm/month; 95% CI: 0.027, 0.092; p = 0.001] and weight (diff: 0.031 kg/month; 95% CI: 0.019, 0.042; p < 0.001). The monthly changes in WAZ (diff: 0.028 z score/month; 95% CI: 0.016, 0.039) and WLZ (diff: 0.042 z score/month; 95% CI: 0.024, 0.059) were also found to be slower, favoring children in the intervention area, compared with the control area (all Ps < 0.001).

NR, non-reported; I, intervention; C, comparison.