Table 3 Summary of included studies

Ahmed et al. [24] [Bangladesh]

Randomized controlled trial

The Transfer Modality Research Initiative (TMRI) provided:

Cash or food transfers, with or without nutrition behavior change communication (BCC) for rural women living in poverty

All four interventions (Cash transfer, food transfer, cash + food, cash + BCC) increased monthly per capita food consumption, daily per capita intake caloric, and food consumption score. The effects are slightly higher for cash + BCC, particularly from the food consumption score

Bandiera et al. [25] [Bangladesh]

Randomized controlled trial

The Targeting the Ultra-Poor program provided:

(a) livestock assets and skills transfers for the poorest women. Women were offered a menu of assets to support income generating activities. Assets included livestock and goods for small-scale retail operations, tree nurseries and vegetable growing

(b) each asset was offered with a package of complementary training and support

Food security among women improved, but there was no effect on mental health status

Blakstad et al. [26] [Tanzania]

Randomized controlled trial

The Homestead Food Production Program provided:

(a) agricultural training for women and inputs to promote homestead food production

(b) nutrition and public health counseling for women to improve diet and health-related behaviors

Household dietary diversity score increased, but there was no effect on food security

Bonuedi et al. [27] [Sierra Leone]

Quasi-experimental design

The Pro-Resilience Action (PROACT) project provided:

(a) the LANN was a participatory community-based intervention involving nutrition education, behavioral change communication and awareness creation on the benefits of consuming diverse diets, proper child feeding and water, sanitation, and hygiene (WASH) practices, and sustainable agriculture and natural resource management in rural areas

(b) a cash crop, income-oriented intervention aimed at enhancing economic access to nutritious foods. It included a nutrition program directed at improving nutrition knowledge and stimulating nutrition-sensitive spending and allocation of other household resources

LANN, combined with the cash crop intervention, improved dietary diversity and food consumption among women and children. LANN alone did not have any effect

Deninger et al. [28] [India]

Quasi-experimental design

The District Poverty Initiative in India:

Supported new self-help groups for women living in poverty in India by training leaders and accountants from new self-help groups in basic management and accounting. The SHGs combined savings generation and micro-lending with social mobilization

The creation of SHGs had mixed effects on food consumption (RS/year), energy intake per capita (kcal/day), and protein intake p.c. (g/day) among poor, non-poor, and poorest of the poor. The three outcomes improved among the poor. Energy intake increased also for the poorest of the poor, but the other two outcomes were not significant for them. None of the outcomes improved among the non-poor

Emran et al. [29] [Bangladesh]

Quasi-experimental design

The Targeting the Ultra-Poor (TUP) program provided:

(a) health, education, and training for poor women, including trainings in livestock and poultry rearing; fruit, vegetable, and herb cultivation; operation of tree nurseries; and village vending

(b) vitamin A supplements for children under five

The probability of having two meals a day, the probability of having sufficient food to meet the household's needs, and grain stock increased. The highest impacts were reported on the first two outcomes

Haque et al. [30] [Bangladesh]

Randomized controlled trial

The Suchana project provided:

Training on agriculture, aquaculture, and market development, including challenging the gender barriers to agriculture, health and nutrition practices among the beneficiary women, husbands, and other household members

The Suchana project increased food consumption during pregnancy, the consumption of vitamin A capsules after last delivery, and the consumption of at least 100 IFA tablets during pregnancy. Greater impacts were reported for the first two outcomes

Heckert et al. [31] [Burkina Faso]

Randomized controlled trial

The Enhanced Homestead Food Production (E-HFP) program provided:

(a) agricultural assets

(b) behavior change communication on agricultural activities, optimal infant and young child feeding, health, hygiene, and care practices

The E-HFP program reduced wasting but had null effects on hemoglobin levels among children

Marquis et al. [32] [Ghana]

Quasi-experimental design

The Enhancing Child Nutrition through Animal Source Food Management (ENAM) program provided:

(a) microcredit loans

(b) weekly nutrition, technical, and entrepreneurship training on viable income-generation activities

The program had a positive effect on height-for-age z score, a negative effect on weight-for-age z score, and a null effect on BMI-for-age z score of preschool-aged children

Pan et al. [33] [Uganda]

Quasi-experimental design

This large-scale agricultural extension program for smallholder women farmers provided:

(a) training through model farmers

(b) easier access to and affordability of seeds sold through farmers serving as community agriculture promoters

The program reduced meals skipped, worries about insufficient food, and limited variety of food among smallholder women. It increased per capita food consumption