Growth faltering is widespread in many low- and middle-income countries, but its effects on childhood bone mass accrual are unknown. The objective of this study was to estimate associations between length (conditional length-for-age z-scores, cLAZ) and weight (conditional weight-for-age z-scores, cWAZ) gain in three age intervals (ages 0–6, 6–12 and 12–24 months) with dual-energy X-ray absorptiometry-derived measures of bone mass (total body less head (TBLH) bone mineral content (BMC), areal bone mineral density (aBMD) and bone area) at 4 years of age.

Associations between interval-specific growth parameters (cLAZ and cWAZ) and bone outcomes were estimated using linear regression models, adjusted for maternal, child and household characteristics.

Data collection occurred in Dhaka, Bangladesh.

599 healthy children enrolled in the BONe and mUScle Health in Kids Study.

cLAZ in each age interval was positively associated with TBLH BMC, aBMD and bone area at 4 years; however, associations attenuated towards null upon adjustment for concurrent height-for-age z-scores (HAZ) at age 4 years and confounders. cWAZ from 0 to 6 and 6 to 12 months was not associated with bone mass, but every sd increase in cWAZ between 12 and 24 months was associated with greater BMC (7·6 g; 95 % CI: 3·2, 12·0) and aBMD (0·008 g/cm2; 95 % CI: 0·003, 0·014) after adjusting for concurrent WAZ, HAZ and confounders.

Associations of linear growth (birth to 2 years) with bone mass at age 4 years were explained by concurrent HAZ. Weight gain in the second year of life may increase bone mass independently of linear growth in settings where growth faltering is common.