Weight- (WAZ), height- (HAZ), and BMI-for-age (BMIZ) are frequently used to assess malnutrition among children. These measures represent different categories of risk and are usually hypothesized to be affected by distinct factors, despite their inherent relatedness. Life history theory suggests weight should be sacrificed before height, indicating a demonstrable relationship among them. Here we evaluate impact of family composition and household economics on these measures of nutritional status and explore the role of WAZ as a factor in HAZ. Anthropometrics, family (...) demographics, and measures of household economy were collected from Sidama agropastoralist children in a peri-urban village in southwestern Ethiopia (n = 157; 79 girls). Just over half of the sample (50.9%) had z-scores of − 2SD or below on at least one measure, indicating an elevated risk of morbidity/mortality; 30% were at or below − 2SD on two or more measures. We used hierarchical linear regression with random intercept analysis to model WAZ and HAZ. Siblings and crop sales significantly decrease WAZ while electricity, agriculture, and polygyny improve z-scores; however, an interaction between polygyny and siblings indicates negative effects of siblings in polygynous families and positive effects in nonpolygynous ones (adj. R2 = 66.5%). For HAZ, agriculture and electricity are positively associated with z-scores whereas siblings have a negative effect; the interaction term again indicates that effects of siblings vary in polygynous and nonpolygynous families (adj. R2 = 74.2%). A mediation model exploring the role of weight in height outcomes suggests not only that WAZ has direct effects on HAZ but also that effects of electricity and agriculture on HAZ are partially mediated by WAZ. Our findings indicate that WAZ and HAZ are primarily affected by shared variables, but effects of siblings vary by polygyny status. Long-term outcomes (HAZ) among Sidama children would likely benefit from interventions focused on stabilizing WAZ across family members. (shrink)