|Timing of growth stunting differs between indigenous Guatemalan boys and girls and maternal education attenuates stunting in girls
|Tumilowicz, Torres, Alison Therese; Pelletier, David; Habicht, Jean Pierre; Pelto, Gretel
|FASEB Journal, 2006; 20(5, Part 2): A1052
Previous studies of endemic undernutrition in Guatemalan infants and young children have not identified gender differences in stunting. However, in ethnographic research in an indigenous community, I discovered differences in mothers' perceptions of the nutritional needs of male and female children. To test whether these perceptions had the expected influence on growth, I used the 1999 Guatemalan Demographic and Health Survey, separating out the indigenous sample. Although attained size was similar at 36 mos., the early trajectory of stunting in boys was steeper than in girls, resulting in a mean difference of 0.45 Z-score (p<.005) by 12 mos. of age. After 14 months, faltering in males leveled off and females continues to decline. Neither magnitude nor statistical significance of the association of gender, age and growth is due to confounding by SES or other factors we examined. Regression models show that many predictors of growth are different (p<.05) for male and female children, and even those variables that are statistically associated with growth in both have different relationships to growth. For example, son preference by mothers has an effect on growth for both genders. The detrimental effect on girls of son preference is ameliorated by a combination of mother's education and improved flooring in the household. Although son preference is beneficial to boys in most circumstances, boys with mothers that have a son preference and need translation between an indigenous language and Spanish are among the most stunted in the sample. The evidence that social predictors of growth faltering are different for boys and girls needs attention from planners so that policies and programs can be more effectively designed in relation to the social realities that they are intended to address.