Mother’s decision making autonomy and grandparent(s) in the household Mother’s high decision-making autonomy was 77.2% in the study population and 81.5% among mothers of children excluded from this analysis. Roughly half of children from both the study population (54.3%) and those excluded (54.7%) lived with paternal grandparent(s) in the household, where 38.3% and 41.9% of household heads were a grandparent, respectively. High decision-making autonomy was higher among mothers who lived in households with no grandparents present (84%) than mothers who lived in households with any type of grandparent (e.g., 67%-74% for mothers who lived with a child’s grandparent) (Table 2). Table 3 presents the unadjusted and adjusted associations of maternal decision-making autonomy, the presence of grandparents in the household as (1) co-residents of the household and (2) head of the household, as well as other socio-demographic factors with child XXX. In the unadjusted analysis, we observed no statically significant association between high decision- making autonomy and child HAZ in comparison to low decision-making autonomy (unadjusted: -0.08 [-0.25, 0.095]; adjusted: 0.04 [-0.21, 0.14]). Relative to children with no grandparents in the home, children with paternal grandparent(s) only in the household had 0.33 SD higher height-for-age. This association remained positive but was weaker and no longer statistically significant in the adjusted model, suggesting that the value of grandparents may be mediated through the socio-demographic variables. No other types of grandparents were statistically significantly associated with child HAZ. Children from households where a grandparent was the household head had higher HAZ scores (0.36, 95% CI: 0.22, 0.49) in comparison to children from households where their father was the head. This association was positive but not statistically significant in the adjusted model (0.10, 95% CI: -0.15, 0.35). When individual maternal autonomy indicators were observed separately in relation to HAZ (see Supplemental Table 2), only action taken when the mother herself falls sick was significantly, but negatively, associated with height-for-age in the unadjusted models (-0.13, 95% CI: -0.25, -0.01). When adjusted for all other variables and covariates, the indicator was still negatively and significantly associated with height-for-age (-0.34, 95% CI: -0.50, -0.17). Two other indicators were also significant, but positively: decisions to buy land or property (0.25, 9...
