Andersen 2015.

Study characteristics
Methods	Study design: PCS How were missing data handled: authors excluded participants without complete covariate and anthropometric outcome data from the analysis. Randomisation ratio: N/A Recruitment method: Peruvian sample recruited from 20 sampling sites selected to reflect diversity in region, ethnicity and religion; however, authors did not report how recruitment was done. Sample size justification and outcome used: NR Sampling method: within the study sites, children within the eligible age category (6–18 months) were randomly sampled for participation. Study aim or objective: to estimate the association of participation in Peru's Juntos CCT with anthropometry, language development and school achievement among children aged 7–8 years. Study period: initial recruitment of 6‐ to 18‐month‐old children started in 2002 with interim follow‐up data collected in 2006 (children aged 4–6 years) and final follow‐up data collected in 2009 (children aged 7–8 years). Unit of allocation or exposure: HHs
Participants	Baseline characteristics Intervention or exposure group (n = 374) Age: mean, months: intervention > 2 years: all (n = 169): 11.3 (SD 3.57); female (n = 84): 11.6 (SD 3.70); male (n = 85): 11.0 (SD 3.43). Participated in intervention ≤ 2 years: all (n = 188): 11.8 (SD 3.54); female (n = 100): 12.0 (SD 3.39); male (n = 88): 11.7 (SD 3.73). Mean percentage of HH aged 0–5 years: intervention ≤ 2 years: total group 30.7 (SD 12.6) people, females 31.3 (SD 12.0), males 30.1 (SD 13.4); intervention > 2 years: total group 32.6 (SD 13.2), females 32.3 (SD 13.7), males 32.9 (SD 12.6). Mean percentage of HH aged 6–14 years: intervention ≤ 2 years: total group 20.2 (SD 18.0), females 19.6 (SD 17.8), males 21.0 (SD 18.4); intervention > 2 years: total group 18.9 (SD 17.3), females 17.2 (SD 16.9), males 20.6 (SD 17.6). Place of residence: HH in rural area: intervention ≤ 2 years: total group 145/188 (77.1%), females 81/100 (81.0%), males 64/88 (72.7%); intervention > 2 years: total group 150/169 (88.8%), females 74/84 (88.1%), males 76/85 (89.4%). Sex: females, n (%): intervention ≤ 2 years: 100/188 (53.2); intervention > 2 years: 84/169 (49.7) Ethnicity and language: carer's first language was indigenous, n (%) – participating in Juntos > 2 years – all: 152 (89.9); females: 82 (97.6); males: 70 (82.4). participating in Juntos ≤ 2 years: all: 130 (69.1); females: 66 (66.0); males: 64 (72.7) Occupation: NR Education: carer completed primary education, n (%): intervention ≤ 2 years: total group 73/188 (38.8), females 40/100 (40.0), males 33/88 (37.5); intervention > 2 years: total group 36/169 (21.3), females 16/84 (19.0), males 20/85 (23.5). TVIP, mean (SD): total group –0.72 (SD 0.96), females –0.82 (SD 0.97), males –0.62 (SD 0.94). SES: mean HH wealth index: intervention ≤ 2 years: total group 0.28 (SD 0.10), females 0.27 (SD 0.10), males 0.29 (SD 0.10); intervention > 2 years: total group 0.23 (SD 0.10), females 0.23 (SD 0.10), males 0.23 (SD 0.09). Mean number of HH members: intervention ≤ 2 years: total group 6.06 (SD 2.41), females 5.97 (SD 2.43), males 6.17 (SD 2.40); intervention > 2 years: total group 5.91 (SD 2.23), females 5.92 (SD 2.32), males 5.89 (SD 2.16). Social capital: NR Nutritional status: stunting (HAZ < –2SD), n (%) – participating in Juntos > 2 years: all 101 (SD 59.8); females 47 (SD 56.0); males 54 (SD 63.5). Participating in Juntos ≤ 2 years: all 91 (SD 48.4); females 36 (SD 36.0); males 55 (SD 62.5). Overweight (BMIZ > 1), n (%) – participating in Juntos > 2 years: all 65 (38.5); females 38 (45.2); males 27 (31.8). Participating in Juntos ≤ 2 years: all 65 (34.6); females 33 (33.0); males 32 (36.4) Morbidities: NR Concomitant or previous care: NR Control (n = 586) Age: mean, months: total: 11.6 (SD 3.50); females 11.6 (SD 3.50); males 11.6 (SD 3.52). Mean percentage of HH aged 0–5 years: total group 30.1 (SD 12.8), females 30.7 (SD 13.2), males 29.6 (SD 12.3). Mean percentage of HH aged 6–14 years: total group 15.9 (SD 16.4), females 15.2 (SD 16.1), males 16.6 (SD 16.7). Place of residence: HH in rural area: total group 156/557 (28.0%), females 81/274 (29.6%), males 75/283 (26.5%). Sex, n (%) of females: participated in Juntos > 2 years: 84/169 (49.7); participated in Juntos ≤ 2 years: 100/188 (53.2) Ethnicity and language: carer's first language was indigenous, n (%): all 180/557 (32.3); females 90/274 (32.8); males 90/283 (31.8) Occupation: NR Education: carer completed primary education, n (%): all (n = 557) 408 (73.2); female 202 (73.7); male 206 (72.8). TVIP mean score (SD): total 0.031 (0.98), females 0.0039 (1.0), males 0.058 (0.96) SES: HH wealth index, mean: all 0.44 (SD 0.17); females 0.43 (SD 0.17); males 0.45 (SD 0.17). Mean number of HH members: total group 5.66 (SD 2.20), females 5.70 (SD 2.22), males 5.63 (SD 2.18). Social capital: NR Nutritional status: stunting (HAZ –2SD), n (%): all 170 (30.5); females 78 (28.5); males 92 (32.5). Overweight (BMIZ > 1), n (%). all 247 (44.3); females 127 (46.4); males 120 (42.4) Morbidities: NR Concomitant or previous care: NR Overall: NR Inclusion criteria: children from the Peruvian section of the Young Lives study (poorer districts); from mountain regions only; with full Juntos participation data; from the younger cohort (aged 6–18 months at recruitment) of the Young Lives study; full covariate and anthropometric data for the 3 rounds; having had round 2 receptive vocabulary assessments completed before recruitment (if any) into the intervention; full covariate data as well as language development and school achievement outcomes at final follow‐up. Exclusion criteria: none reported Baseline differences: significant differences between intervention recipients and non‐recipients for nearly all covariates at round 1, all of which indicated an increased level of vulnerability and poverty among intervention participants; e.g. more likely to live in rural areas, have a lower wealth index, have a carer who spoke an indigenous language, and have a carer who did not complete primary education. Total number enrolled per relevant group: anthropometric: intervention > 2 years = 179 children; intervention ≤ 2 years = 195 children; controls = 586 children. Language development/school achievement: intervention = 272; controls = 586. Total number randomised per relevant group: N/A Total number completed and analysed per relevant group: anthropometric outcomes: intervention > 2 years: 188; girls 100; boys 88. Intervention ≤ 2 years: 169; girls 84; boys 85. Language development/school achievement: intervention = 243 children, control = 521 children. Attrition per relevant group: intervention (anthropometric outcomes) = 17 children (7/195 exposed for ≤ 2 years and 10/179 exposed for > 2 years); control (anthropometric outcomes) = 29/586 children; intervention (language development/school achievement outcomes) = 29/272 children; control (language development/school achievement) = 65/586 children. No reasons for attrition provided. Description of subgroups measured and reported: intervention group divided into 2 subgroups for anthropometric analysis: intervention for ≤ 2 years and Intervention for > 2 years. Both groups were compared with unexposed controls. Results presented for girls and boys.
Interventions	Intervention characteristics Intervention or exposure Food access intervention category: increase buying power Intervention type: CCT programme Description: eligible HHs received cash transfer from government; beneficiary HHs received transfers of PEN 100 (USD 30) each month regardless of HH composition, representing, about 15% of beneficiary HH spending. Conditionalities: members of HHs with children aged 5 years or with a pregnant or lactating woman were required to attend regular healthcare visits. Children aged 6–14 years who had not completed primary school were required to attend school 85% of the days. Duration of intervention period: up to 5 years (from Juntos inception in 2005 to Young Lives round 3 follow‐up in 2009), distinguished as ≤ 2 years and > 2 years Frequency: monthly CCT Number of study contacts: baseline (2002) with 2 follow‐ups (2006 and 2009) Providers: government Delivery: NR Co‐interventions: NR Resource requirements: NR Economic indicators: intervention cost ≥ PEN 100 per month per HH Control: no intervention
Outcomes	Anthropometry: HAZ, stunting, BMI‐for‐age Cognitive function and development: language (TVIP) score, grade attainment Adverse event: overweight
Identification	Sponsorship source: Bill Melinda Gates Foundation (Global Health grant OPP10327313), the Eunice Kennedy Shriver National Institute of Child Health and Development (grant R01 HD070993) and Grand Challenges Canada (grant 0072‐03 to the grantee, the Trustees of the University of Pennsylvania). The Young Lives Study was core funded by the UK Aid from the DfID and cofunded from 2010 to 2014 by the Netherlands Ministry of Foreign Affairs. Country: Peru Setting: poor HHs in poor districts Authors' names: Christopher T Andersen; Lia CH Fernald Email: chrisandersen@berkeley.edu; fernald@berkeley.edu Type of record: journal article Declarations of interest: no conflicts of interest. Study or programme name and acronym: Young Lives Study; Juntos conditional cash transfer programme
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (Selection bias)	High risk	Cohort study and no randomisation performed.
Allocation concealment (Selection bias)	High risk	Cohort study and no allocation concealment performed.
Baseline characteristics similar (Selection bias)	Low risk	Baseline characteristics balanced by PSM.
Baseline outcome measurements similar (Selection bias)	High risk	Serious baseline imbalances for overweight, stunting and TVIP score outcomes that were not adjusted for when matching controls to participants. Overweight participants were significantly lower (P < 0.05); and stunting was significantly higher (P < 0.01) for intervention participants. TVIP scores were significantly lower (P < 0.01) among intervention participants.
Blinding of participants and personnel (Performance bias)	Low risk	Cohort study and no blinding performed. However, this was unlikely to affect objective outcomes of weight and height.
Blinding of outcome assessment (Detection bias)	Low risk	Unclear whether study staff assessing outcomes were aware of Juntos exposure during the assessment procedure; however, outcomes were objective.
Protection against contamination (Performance bias)	Unclear risk	Unclear whether control participants may have benefited indirectly from Juntos through eligible HHs in their community (e.g. a control child taking meals at his/her friend's participating home).
Incomplete outcome data (Attrition bias)	Low risk	Although children without complete outcomes data were excluded from the analysis, similar proportions were excluded from the control group (n = 29, 4.9%) and from intervention groups (n = 10, 5.5% and n = 7, 3.6%), and outcomes were frequent enough that it was unlikely that the small numbers missing would greatly change the effect observed. In the Young Lives sample less than (quote) "3% of children were completely lost to follow‐up between rounds 1 and 3. Those lost to follow up were more likely to have a caretaker who spoke an indigenous language, but they were similar across all other covariates and baseline outcomes."
Selective outcome reporting (Reporting bias)	Unclear risk	No protocol available. All a priori stated outcomes in the Methods section were reported in the Results section.
Other bias	Low risk	None identified.