Fig. 1

Birth weight changes associated with per 3% increase in energy from dietary protein intake during pregnancy. Multilevel linear regression models were used to estimate changes and 95% CIs. Model 1 was adjusted for total energy intake, offspring sex, gestational age, and socio-demographic characteristics (including geographic area, residence, childbearing age, education, occupation, household wealth index, and parity). Model 2 was adjusted for all variables in Model 1 plus health-related characteristics (including passive smoking, alcohol drinking, antenatal check visit frequency, folate/iron supplements use, anemia, and medication use) and principal component score based on the nutrient intakes. Models were mutually adjusted for animal protein and plant protein. The black circles represent birth weight changes, and the vertical lines represent 95% CIs