Table 2 Effects of the disability-inclusive graduation programme across timepoints. Intervention effects were estimated using linear mixed-effects regression, reporting minimally adjusted mean differences (MAMDs) and fully adjusted mean differences (FAMDs) with 95% confidence intervals (CIs). The minimally adjusted model included treatment status (fixed effect) and cluster/branch (random intercepts). Fully adjusted models additionally controlled for imbalanced variables (p < 0.10): marital status and age of project participants for first follow-up, and these variables plus household per capita income for second follow-up

First follow-up (0 month after the intervention)

 Social participation

3.04 (24.83)

 − 0.29 (26.92)

3.04 [− 1.98, 8.04]

0.24

0.12 [− 0.08, 0.31]

3.14 [− 1.26, 7.54]

0.17

0.14 [− 0.06, 0.34]

Domain 1: Household participation

 − 0.8 (28.4)

 − 0.95 (28.81)

0.01 [− 4.92, 4.89]

1.00

0 [− 0.17, 0.17]

 − 0.24 [− 4.49, 3.98]

0.91

 − 0.01 [− 0.19, 0.17]

Domain 2: Societal participation

5.92 (33.65)

0.21 (36.08)

5.49 [− 1.59, 12.55]

0.13

0.16 [− 0.05, 0.36]

5.84 [− 0.01, 11.69]

0.05

0.20 [0, 0.41]

Second follow-up (16 months after the intervention)

 Social participation

2.41 (25.2)

 − 0.01 (27.77)

2.25 [− 2.83, 7.28]

0.38

0.08 [− 0.11, 0.28]

2.56 [− 1.79, 6.90]

0.25

0.11 [− 0.08, 0.31]

Domain 1: Household participation

 − 2.67 (28.04)

 − 1.85 (27.1)

 − 0.82 [− 5.45, 3.81]

0.73

 − 0.03 [− 0.2, 0.14]

 − 0.22 [− 4.12, 3.69]

0.91

 − 0.01 [− 0.18, 0.16]

Domain 2: Societal participation

6.23 (35.37)

1.38 (37.44)

4.68 [− 2.56, 11.90]

0.20

0.13 [− 0.07, 0.33]

4.85 [− 1.02, 10.72]

0.11

0.17 [− 0.04, 0.37]