The association between paternal labour migration and the growth of the left-behind children-evidence from a birth cohort in Dhanusha district, Nepal.
Laura Busert-Sebela, Mario Cortina-Borja, Jonathan Wells, Delan Devakumar, Simon Eaton, Dharma S Manandhar, Shyam Sundar Yadav, Naomi M Saville
Abstract
Open AccessINTRODUCTION: We aimed to determine the association between paternal labour migration and the growth of the left-behind children in Dhanusha district, Nepal, where child stunting and international labour migration are highly prevalent. METHODS: We used growth data at birth, 6 months, 1 year and 2 years from a birth cohort study conducted 2012-2014, and growth data at age 6 years collected in 2018. We collected household migration history data to determine the children's exposure to paternal migration. The primary outcome was child length/height-for-age z-score (HAZ). Children's body circumferences, skinfold thicknesses, body composition, tibia length and grip strength were secondary outcomes measured at 6 years. We tested (i) the overall association between paternal international migration and the growth of the left-behind child; the roles of (ii) the duration of migration (≤12 mvs >12 m) and (iii) child age (≤6 mvs 12-72 m) as moderating factors; (iv) the association between receipt of remittances from the migrant father and child growth outcomes; and (v) stratified the main analyses by child gender. We fitted mixed-effects linear regression models for longitudinal data and linear regression models for cross-sectional data, adjusted for potential confounders. RESULTS: Analysing across all time points, daughters of labour migrants had lower HAZ than daughters of non-migrants (-0.13, 95% CI -0.24 to -0.03), but no overall association was found in boys. The negative associations were largest at <6 m (girls: -0.23, 95% CI -0.41 to -0.05), but in boys only if the father had recently (≤12 m) migrated (-0.26, 95% CI -0.51 to 0.00). Children of migrants showed a tendency towards smaller body sizes compared with children of non-migrants. We found no association between remittances and any measure of child growth. CONCLUSIONS: Interventions should target support for pregnant women and mothers with young infants to provide gender-equitable childcare, especially if their husband just left for work overseas.