Table 4.
Effect of interventions on other outcomes adjusted for baseline measures of the outcomes and stratification factors (caste and distance from nearest town)
| Control | AGRI | AGRI-NUT | AGRI-NUT+PLA |
AGRI vs control |
AGRI-NUT vs control |
AGRI-NUT+PLA vs control |
||||
|---|---|---|---|---|---|---|---|---|---|---|
| Adjusted relative risk (95% CI) | p value | Adjusted relative risk (95% CI) | p value | Adjusted relative risk (95% CI) | p value | |||||
| Maternal low MUAC <230 mm | 445/989 (45%) | 539/1096 (49%) | 465/1051 (44%) | 525/1134 (46%) | 1·08 (0·96 to 1·23) | 0·20 | 0·96 (0·85 to 1·08) | 0·50 | 1·01 (0·90 to 1·13) | 0·85 |
| Child low MUAC <125 mm* | 96/783 (12%) | 128/861 (15%) | 104/827 (13%) | 110/893 (12%) | 1·16 (0·85 to 1·57) | 0·35 | 1·00 (0·76 to 1·31) | 0·98 | 0·92 (0·66 to 1·28) | 0·62 |
| Maternal haemoglobin (g/dL)† | 11·7 (1·30), n=974 | 11·6 (1·29), n=1077 | 11·6 (1·30), n=1033 | 11·6 (1·29), n=1111 | 0·02 (−0·15 to 0·19) | 0·84 | 0·00 (−0·17 to 0·17) | 0·98 | −0·13 (−0·30 to 0·04) | 0·15 |
| Child haemoglobin (g/dL)* | 10·1 (1·19), n=776 | 10·2 (1·29), n=857 | 10·1 (1·26), n=825 | 10·1 (1·25), n=883 | 0·11 (−0·07 to 0·28) | 0·23 | 0·04 (−0·13 to 0·21) | 0·68 | −0·05 (−0·23 to 0·14) | 0·62 |
| Child minimum acceptable diet* | 268/790 (34%) | 290/862 (34%) | 327/829 (39%) | 384/895 (43%) | 1·01 (0·86 to 1·19) | 0·89 | 1·19 (1·02 to 1·41) | 0·032 | 1·30 (1·12 to 1·52) | 0·0008 |
| Women made ≥2 decisions in agriculture or health | 867/997 (87%) | 1015/1100 (92%) | 961/1055 (91%) | 1002/1139 (88%) | 1·05 (1·00 to 1·11) | 0·038 | 1·05 (0·99 to 1·1) | 0·12 | 1·02 (0·96 to 1·07) | 0·55 |
| Women worked <10·5 h in the previous 24 h | 339/997 (34%) | 339/1100 (31%) | 332/1055 (31%) | 357/1139 (31%) | 0·79 (0·60 to 1·05) | 0·11 | 0·84 (0·62 to 1·15) | 0·28 | 0·76 (0·56 to 1·03) | 0·082 |
| Women achieving gender parity in agriculture‡ | 253/406 (62%) | 283/439 (64%) | 264/419 (63%) | 294/465 (63%) | 1·02 (0·91 to 1·13) | 0·77 | 1·01 (0·90 to 1·13) | 0·90 | 0·98 (0·87 to 1·10) | 0·72 |
| Share of household expenditures spent on food§ | 0·62 (0·17) n=500 | 0·63 (0·17) n=546 | 0·62 (0·17) n=531 | 0·63 (0·17) n=569 | 0·00 (−0·04 to 0·03) | 0·80 | −0·01 (−0·04 to 0·03) | 0·69 | 0·01 (−0·02 to 0·04) | 0·66 |
| Per capita total daily household expenditure (INR)§¶ | 17·6 (12·9 to 25·7), n=500 | 18·2 (13·5 to 26·6), n=546 | 18·2 (13·4 to 26·7), n=531 | 18·2 (12·6 to 26·1), n=569 | 1·07 (−1·70 to 3·56) | 0·43 | 0·83 (1·57 to 3·20) | 0·50 | 2·39 (−0·35 to 5·40) | 0·096 |
| Agricultural production diversity out of 10 food groups over 1 year | 4·1 (2·03), n=996 | 4·4 (2·03), n=1100 | 4·6 (1·95), n=1053 | 4·4 (2·02), n=1138 | 0·13 (−0·25 to 0·50) | 0·51 | 0·37 (0·00 to 0·75) | 0·052 | 0·31 (−0·07 to 0·69) | 0·11 |
| Total value of agricultural production over 1 year (INR)¶ | 16 696 (6312 to 34 400), n=996 | 19 465 (8475 to 37 648), n=1100 | 19 668 (9229 to 38 368), n=1053 | 17 378 (7395 to 34 683), n=1138 | 7579 (2211 to 16 298) | 0·03 | −323 (−9442 to 8295) | 0·94 | 4922 (−163 to 13 407) | 0·15 |
| Net value (total value minus input costs) of agriculture production over 1 year (INR)¶ | 10 048 (2810 to 24 241), n=996 | 13 489 (4405 to 28 118), n=1100 | 13 581 (4792 to 27 869), n=1053 | 11 885 (3528 to 25 779), n=1138 | 6825 (1781 to 15 412) | 0·045 | −197 (−9218 to 7909) | 0·96 | 5324 (857 to 13 817) | 0·11 |
Data are mean (SD), n/N (%), or median (IQR), unless indicated otherwise. All denominators are valid responses from the endline survey. NSA=nutrition-sensitive agriculture. AGRI=group assigned to NSA videos. AGRI-NUT=group assigned to NSA and nutrition-specific videos. AGRI-NUT+PLA=group assigned to NSA videos and a nutrition-specific participatory learning and action cycle meetings and videos. MUAC=mid-upper arm circumference.
Includes children aged 6–23 months only.
Includes non-pregnant women only.
Gender parity achieved when women have equal or higher empowerment scores than men; empowerment scores are calculated for men and women as weighted sums of five indicators: decision making, asset ownership, access to credit, group membership, and time use; measured in 50% of households, excludes female-only households.
Measured in 50% of households.
Confidence intervals are bias corrected estimated using non-parametric bootstrapping.