William Brubaker and Dr. Kirk Dearden, Health Science

Children experience extreme poverty differently than adults. Child poverty cannot only be categorized in terms of family income since it is experienced as both material and developmental deprivation. In order to fully assess the effects of childhood poverty, children’s experiences based on set outcomes must be taken into account. We used a dataset collected by Young Lives, a long-term longitudinal study conducted in India, Ethiopia, Vietnam and Peru investigating the changing nature of childhood poverty. In our proposed study we used a cohort of 1,000 8-year-olds from each country to examine the relationship between childhood unwantedness and health, educational, and mental outcomes.

We started out by constructing a conceptual framework to identify and select a few proxy indicators for child unwantedness and looked for possible associations with these variables and other selected variables from our framework. One of the first problems was finding universal variables that were used in each of the surveyed countries. Of the four countries, Peru contained the best proxy indicator variables including whether the mother wanted to become pregnant when she became pregnant with the index child and whether she breastfed the child. With these variables we performed bivariate analysis followed by a logistic regression model. Unfortunately, the models were severely underpowered due to small sample size (low levels of unwantedness) and we were unable to produce any noteworthy results.

The inconclusive results due to sample size were caused initially by the difficulty of accurately defining child unwantedness with proxy variables from the given dataset. Residual confounders may have also played a role in the inability to attain significant results. Additional study in the area of child unwantedness is warranted but with the use of specific predictor variables which would clearly point to significant outcomes.

We decided to adjust the original conceptual framework to examine nutritional status of children beginning with Peru alone. Using the same Young Lives dataset, we have looked at predictors of two dichotomous outcomes involving positive deviants. The first compared positive deviants (those with good nutritional outcomes and poor economic indicators) against negative deviants (those with poor nutritional outcomes and good economic indicators). The results revealed that only the sex of child, access to food, and the mother joining a community group to address a problem or issue were significant predictors. The second dichotomous outcome, which compared positive deviants to those with poor nutritional outcomes and poor economic indicators, showed that sex, access to food, and whether or not the child worked in the previous 12 months were the only significant predictors.

Although we did identify predictors that were statistically significant, they were all expected and no variable was novel and noteworthy. Further research is warranted and we will continue to run the logistical regression model for the remaining countries. If any significant and noteworthy results are attained we will use them to produce a manuscript that will be submitted to a peer-reviewed publication. The findings will also be presented at an international conference on global health.

Reference

• UNICEF, State of the World’s Children 2005, New York, 2004, p. 16.