14. Income‐generation interventions – results of included trials.
| Study ID (risk of bias) | Study design (n) | Income‐generation interventions | No intervention | Effect measure (time point) | Effect direction | Meta‐analysis (yes/no) | |||||||||||||||
| Results at baseline | Results at follow‐up | n | Results at baseline | Results at follow‐up | n | ||||||||||||||||
| Primary outcomes | |||||||||||||||||||||
| 3.3 Proportion of HHs who were food secure | |||||||||||||||||||||
| 3.3.1 Food security | |||||||||||||||||||||
| 3.3.1.1 Outcome measure: proportion experiencing food security (based on HFIAS) | |||||||||||||||||||||
| Osei 2017 (?) | cRCT (2614 HHs) | 79.7, 95% CI 77.2 to 82.0 | 53.6, 95% CI 51.0 to 56.1 | — | 87.4, 95% CI 85.3 to 89.3 | 78.3, 95% CI 76.0 to 80.4 | — | — (2.5 years) | — | No n for individual groups to calculate MD. | |||||||||||
| 3.3.1.2 Outcome measure: HH food security score (mean, SD) | |||||||||||||||||||||
| Beegle 2017 (‐) | RCT (2193 HHs) | — | — | 1083 HHs | — | –3.12 (1.29) | 1110 HHs | MD –0.060, SE 0.080, 95% CI –0.2168 to 0.0968 (3/4 months) | ▽ | ||||||||||||
| 3.3.1.3 Outcome measure: Resilience Index (mean, SD) | |||||||||||||||||||||
| Beegle 2017 (‐) | RCT (2195 participants) | — | — | — | — | –9.32 (9.84) | — | MD –0.224, SE 0.630 (3/4 months) | ▽ | ||||||||||||
| 3.3.1.4 Outcome measure: Principal Components Analysis index (mean, SD) | |||||||||||||||||||||
| Beegle 2017 (‐) | RCT (2123 participants) | — | — | — | — | 0.15 (2.08) | — | MD –0.029, SE 0.135 (3/4 months) | ▽ | ||||||||||||
| 3.3.2 Dietary diversity | |||||||||||||||||||||
| 3.3.2.1 Outcome measure: odds of consuming an additional food group based on the DDS | |||||||||||||||||||||
|
Darrouzet Nardi 2016 (?) (DDS 0–7) |
cRCT (2584 children) | — | — | — | — | — | — | OR 1.524, 95% CI 1.45 to 4.38, P = 0.001 (2 years) | ▲ | ||||||||||||
| 3.3.2.2 Outcome measure: HDDS (mean, SD)/Food Consumption Score | |||||||||||||||||||||
|
Olney 2016 (?) (HDDS 0–11) |
cRCT (1476 HHs) | 5.6 (1.93) | 5.6 (2.07) | 880 HHs | 5.8 (1.70) | 5.2 (2.11) | 596 HHs | MD 0.7, SE 0.44, 95% CI –0.1624 to 1.5624, P = 0.17 (2 years) | △ | Yes. SMD. | |||||||||||
|
Beegle 2017 (‐) (FCS 0–126) |
RCT (2201 HHs) | — | — | 1191 HHs | — | 38.82 (16.01) | 1110 HHs | MD –0.708, SE 1.072, 95% CI –2.80912 to 1.39312 (3/4 months) | ▽ | ||||||||||||
| 3.3.2.3 Outcome measure: MDD (n, %) | |||||||||||||||||||||
| Marquis 2018 (+) | cRCT (428 children) | 30.9 | 80.2 | 247 | 33.8 | 69.5 | 181 | OR 1.65, SE 0.41, 95% CI 0.8464 to 2.4536, P < 0.05 (12 months) | ▲ | Yes. Olney groups combined. | |||||||||||
| Darrouzet Nardi 2016 (?) | cRCT (2604 children) | — | — | — | — | — | — | OR 1.146, 95% CI 1.02 to 1.29, P = 0.021 (2 years) | ▲ | ||||||||||||
| Olney 2016 (?) | cRCT (758 children) | OWL: 7 (3.0) HC: 4 (1.7) |
OWL: 35 (15.0) HC: 43 (18.2) |
OWL: 220 HC: 231 |
8 (2.6) | 20 (6.3) | 307 | OWL villages vs control: pp 8.3, P = 0.17 HC villages vs control: pp 12.6, P = 0.08 (2 years) Combined effect: MD pp 10.08, 95% CI 1.02 to 19.14 |
▲ | ||||||||||||
| Secondary outcomes | |||||||||||||||||||||
| 3.5Change in anthropometric indicators | |||||||||||||||||||||
| 3.5.1 Stunting | |||||||||||||||||||||
| 3.5.1.1 Outcome measure: Height‐for‐Age z‐score (HAZ) (mean, SD or SE) | |||||||||||||||||||||
| Marquis 2018 (+) | cRCT (428 children) | –0.88 (1.27) | — | 247 | –0.78 (1.30) | — | 181 | MD 0.22, SE 0.06, P < 0.01, 95% CI 0.10 to 0.34 (12 months) | ▲ | No. Effect sizes calculated for Darrouzet (2 years) and Osei from group estimates. | |||||||||||
| Darrouzet Nardi 2016 (?) | cRCT (303 children) | –1.47 (0.07) | –1.38 (0.06) | — | –1.48 (0.06) | –1.41 (0.06) | — | MD 0.109, 95% CI 0.000 to 0.218, P = 0.048 (12 months) | △ | ||||||||||||
| 609 children | –1.47 (0.07) | –1.30 (0.06) | 305 | –1.48 (0.06) | –1.33 (0.06) | 304 | MD 0.03, SE 0.0049, 95% CI 0.020 to 0.040 (2 years) | ▲ | |||||||||||||
| Osei 2017 (?) | cRCT (2569 children) | –2.23 (0.03) | –2.1 (0.03) | 1299 | –2.4 (0.04) | –2.32 (0.03) | 1297 | MD 0.22, SE 0.0012, 95% CI 0.218 to 0.222 (2.5 years) | ▲ | ||||||||||||
| 3.5.1.2 Outcome measure: proportion stunted (HAZ < –2SD) (CI) | |||||||||||||||||||||
| Osei 2017 (?) | cRCT (2569 children) | 57.7 | 55.1 | 1299 | 65.8 | 63.5 | 1297 | OR 0.94, 95% CI 0.74 to 1.19 (2.5 years) | △ | Yes. Verbowski groups combined | |||||||||||
| Verbowski 2018 and aquaculture (?) | cRCT (597 children) | 27.9 | 29.9 | 299 | 29.3 | 32.0 | 298 | MD pp –0.62, P = 0.927 (1.8 years) | MD pp 2.2, 95% CI –5.64 to 10.05 | ▽ | |||||||||||
| Verbowski 2018 (?) | cRCT (598 children) | 22.7 | 28.9 | 300 | 29.3 | 32.0 | 298 | MD pp 3.73, P = 0.453 (1.8 years) | |||||||||||||
| 3.5.2Wasting | |||||||||||||||||||||
| 3.5.2.1 Outcome measure: WHZ (mean, SD or SE) | |||||||||||||||||||||
| Marquis 2018 (+) | cRCT (428 children) | –0.37 (1.08) | — | 247 | –0.31 (1.24) | — | 181 | MD 0.07, SE 0.08, 95% CI –0.087 to 0.227, P > 0.10 (12 months) | △ | No. Effect for Osei calculated from group estimates. | |||||||||||
| Osei 2017 (?) | cRCT (2603 children) | –0.91 (0.03) | –0.85 (0.03) | 1300 | –0.93 (0.03) | –0.71 (0.03) | 1303 | MD –0.14, SE 0.0012, 95% CI –0.142 to –0.138 (2.5 years) | ▼ | ||||||||||||
| 3.5.2.2 Outcome measure: proportion wasted (WHZ < –2SD) | |||||||||||||||||||||
| Osei 2017 (?) | cRCT (2603 children) | 10.6 | 10.5 | 1300 | 10.1 | 9.7 | 1303 | OR 1.03, 95% CI 0.70 to 1.52 (2.5 years) | ▽ | Yes. Verbowski groups combined. | |||||||||||
| Verbowski 2018 and aquaculture (?) | cRCT (597 children) | 6.7 | 10.2 | 299 | 8.3 | 8.9 | 298 | MD pp 2.75, P = 0.424 (22 months) | MD pp 3.19, 95% CI –1.95 to 8.33 | ▽ | |||||||||||
| Verbowski 2018 (?) | cRCT (598 children) | 8.4 | 13.0 | 300 | 8.3 | 8.9 | 298 | MD pp 3.80, P = 0.348 (22 months) | |||||||||||||
| 3.5.3 Underweight | |||||||||||||||||||||
| 3.5.3.1 Outcome measure: Weight‐for‐age z‐score (WAZ) (mean, SD or SE) | |||||||||||||||||||||
| Marquis 2018 (+) | cRCT (428 children) | –0.78 (1.12) | — | 247 | –0.68 (1.27) | — | 181 | MD 0.15, SE 0.07, P < 0.05 (12 months) | ▲ | Yes. Effect estimates calculated using group estimates. | |||||||||||
| Darrouzet Nardi 2016 (?) | cRCT (634 children) | –2.04 (0.07) | –1.97 (0.06) | 301 | –1.94 (0.06) | –1.89 (0.06) | 333 | NR (1 year) | — | ||||||||||||
| –2.04 (0.07) | –1.97 (0.06) | –1.94 (0.06) | –2.07 (0.06) | MD 0.10, 95% CI 0.09 to 0.11 (2 years) | ▲ | ||||||||||||||||
| Osei 2017 (?) | cRCT (2613 children) | –1.87 (0.03) | –1.77 (0.03) | 1306 | –1.97 (0.03) | –1.77 (0.03) | 1307 | MD 0.00, 95% CI –0.00 to 0.00 (2.5 years) | — | ||||||||||||
| 3.5.3.2 Outcome measure: percentage underweight (WAZ < 80% standard/ < –2SD) (includes severe underweight) | |||||||||||||||||||||
| Osei 2017 (?) | cRCT (2613 children) | 43.4 | 41.0 | 1306 | 48.0 | 40.6 | 1307 | OR 1.15, 95% CI 0.91 to 1.46 (2.5 years) | ▽ | Yes. Verbowski groups combined. | |||||||||||
| Verbowski 2018 and aquaculture (?) | cRCT (597 children) | 23.5 | 32.0 | 299 | 23.0 | 28.8 | 298 | MD pp 2.75, P = 0.670 (22 months) | MD pp –1.16, 95% CI –9.02 to 6.70 | △ | |||||||||||
| Verbowski 2018 (?) | cRCT (598 children) | 26.1 | 28.8 | 300 | 23.0 | 28.8 | 298 | MD pp –3.63, P = 0.479 (22 months) | |||||||||||||
| 3.5.3.3 Outcome measure: BMI (kg/m2) (mean, SD or SE) | |||||||||||||||||||||
| Olney 2016 (?) | cRCT (1297 women) | 20.2 (2.22) | 20.7 (2.34) | 787 | 20.6 (2.27) | 21.1 (2.70) | 510 | MD 0.2, 95% CI –0.192 to 0.592, SE 0.20, P = 0.26 (2 years) | △ | Yes. Effect estimate for Osei calculated from group estimates. | |||||||||||
| Osei 2017 (?) | cRCT (2614 mothers) | 19.6 (0.07) | 19.8 (0.05) | 1182 | 20.1 (0.06) | 19.9 (0.05) | 1303 | MD –0.10, 95% CI –0.10 to –0.10 (2.5 years) | ▼ | ||||||||||||
| 3.5.3.4 Proportion of women who were underweight (BMI < 18.5 kg/m2) | |||||||||||||||||||||
| Olney 2016 (?) | cRCT (1297 women) | 23 | 15 | 787 | 15 | 16 | 510 | pp –8.7, P = 0.01 (2 years) | ▲ | No. Verbowski groups combined. | |||||||||||
| Osei 2017 (?) | cRCT (2614 mothers) | 28.2 | 28.6 | 1182 | 17.5 | 19.9 (0.05) | 1303 | OR 0.61, 95% CI 0.46 to 0.82 (2.5 years) | ▲ | ||||||||||||
| Verbowski 2018 and aquaculture (?) | cRCT (541 women) | 14.2 | 9.0 | 270 | 16.6 | 9.4 | 271 | MD pp 1.19, P = 0.920 (22 months) | MD 3.88, 95% CI –4.36 to 12.12 | ▽ | |||||||||||
| Verbowski 2018 (?) | cRCT (541 women) | 13.4 | 13.5 | 270 | 16.6 | 9.4 | 271 | MD pp 4.27, P = 0.347 (22 months) | |||||||||||||
| 3.6 Change in biochemical indicators | |||||||||||||||||||||
| 3.6.1 Mean haemoglobin concentration (children) (mean, SE) | |||||||||||||||||||||
| Osei 2017 (?) | cRCT (2614 children) | 115.3 (0.1) | 114.3 (0.1) | 1307 | 113.6 (0.1) | 110.8 (0.1) | 1307 | MD 3.5, SE 0.0039, 95% CI 3.492 to 3.507 (2.5 years) | ▲ | Yes. Verbowski groups combined. | |||||||||||
| Verbowski 2018 and aquaculture (?) | cRCT (597 children) | 104.5 (13.7) | 108.4 (13.1) | 298 | 105.7 (13.6) | 107.1 (12.9) | 299 | MD 2.54, SE 1.43, P = 0.076 (22 months) | MD 2.48, 95% CI 0.51 to 4.46 | ▲ | |||||||||||
| Verbowski 2018 (?) | cRCT (597 children) | 104.1 (13.8) | 108.0 (12.3) | 298 | 105.7 (13.6) | 107.1 (12.9) | 299 | MD 2.43, SE 1.42, P = 0.088 (22 months) | |||||||||||||
| 3.6.2 Mean haemoglobin concentration (women) (mean, SD or SE) | |||||||||||||||||||||
| Osei 2017 (?) | cRCT (2614 mothers) | 129.3 (0.1) | 126.5 (0.1) | 1307 | 129.6 (0.1) | 121.9 (0.1) | 1307 | MD 4.6, SE 0.0039, 95% CI 4.592 to 4.608 (2.5 years) | ▲ | Yes. Verbowski groups combined. | |||||||||||
| Verbowski 2018 and aquaculture (?) | cRCT (541 women) | 122.4 (12.1) | 122.9 (12.9) | 270 | 121.5 (12.5) | 121.1 (12.1) | 271 | MD 0.49, SE 1.33, P = 0.714 (22 months) | MD –0.07, 95% CI –1.92 to 1.78 | ▽ | |||||||||||
| Verbowski 2018 (?) | cRCT (541 women) | 121.7 (13.7) | 121.0 (11.9) | 270 | 121.5 (12.5) | 121.1 (12.1) | 271 | MD –0.63, SE 1.34, P = 0.637 (22 months) | |||||||||||||
| 3.9 Morbidity | |||||||||||||||||||||
| 3.9.1 Prevalence of anaemia (children) | |||||||||||||||||||||
| Osei 2017 (?) | cRCT (2614 children) | 28.2 | 30.8 | 1307 | 31.6 | 42.5 | 1307 | OR 0.76, 95% CI 0.59 to 0.98 (2.5 years) | ▲ | Yes. Verbowski groups combined. | |||||||||||
| Verbowski 2018 and aquaculture (?) | cRCT (597 children) | 63.1 | 54.3 | 298 | 59.2 | 59.5 | 299 | MD pp –9.74, P = 0.119 (22 months) | MD pp –11.90, 95% CI –20.47 to –3.33 | ▲ | |||||||||||
| Verbowski 2018 (?) | cRCT (597 children) | 65.4 | 52.6 | 298 | 59.2 | 59.5 | 299 | MD pp –14.0, P = 0.023 (22 months) | |||||||||||||
| 3.9.2 Prevalence of anaemia (women) | |||||||||||||||||||||
| Osei 2017 (?) | cRCT (2614 mothers) | 19.6 | 24.6 | 1307 | 21.1 | 35.8 | 1307 | OR 0.62, 95% CI 0.48 to 0.82 (2.5 years) | ▲ | Yes. Verbowski groups combined. | |||||||||||
| Verbowski 2018 and aquaculture (?) | cRCT (541 women) | 38.9 | 35.8 | 270 | 40.4 | 38.7 | 271 | MD pp –1.10, P = 0.865 (22 months) | MD pp 1.34, 95% CI –7.94 to 10.61 | ▽ | |||||||||||
| Verbowski 2018 (?) | cRCT (541 women) | 41.9 | 43.5 | 270 | 40.4 | 38.7 | 271 | MD pp 4.14, P = 0.551 (22 months) | |||||||||||||
aEach triangle represents one study.
(+): low overall risk of bias; (?): unclear overall risk of bias; (‐): high overall risk of bias. ▲ = Favours the intervention, 95% CI excludes 0; △ = Unclear effect potentially favouring the intervention, 95% CI includes zero; ▼ = Favours the control, 95% CI excludes 0; ▽ = Unclear effect potentially favouring the control, 95% CI includes 0.
BMI: body mass index; cRCT: cluster randomised controlled trial; CI: confidence interval; DDS: Dietary Diversity Score; EHFP: enhanced homestead food production; FCS: Food Consumption Score; HAZ: height‐for‐age z‐score; HC: health committee; HDDS: Household Dietary Diversity Score; HFIAS: Household Food Insecurity Score; HH: household; MD: mean difference; MDD: Minimum Dietary Diversity; n: number; NR: not reported; OR: odds ratio; OWL: older women leaders; pp: percentage point; SD: standard deviation; SE: standard error; SMD: standardised mean difference; WAZ: weight‐for‐age z‐score; WHZ: weight‐for‐height z‐score.