Limits...
Structured additive regression models with spatial correlation to estimate under-five mortality risk factors in Ethiopia.

Ayele DG, Zewotir TT, Mwambi HG - BMC Public Health (2015)

Bottom Line: With respect to socio-economic factors, the greater the household wealth, the lower the mortality.The model enables simultaneous modeling of possible nonlinear effects of covariates, spatial correlation and heterogeneity.Our findings are relevant because the identified risk factors can be used to provide priority areas for intervention activities by the government to combat under-five mortality in Ethiopia.

View Article: PubMed Central - PubMed

Affiliation: School of Mathematics, Statistics and Computer Science, University of KwaZulu-Natal, Private Bag X01, Pietermaritzburg, Scottsville, 3209, South Africa. ayele@ukzn.ac.za.

ABSTRACT

Background: The risk of a child dying before reaching five years of age is highest in Sub-Saharan African countries. But Child mortality rates have shown substantial decline in Ethiopia. It is important to identify factors affecting under-five mortality.

Methods: A structured additive logistic regression model which accounts the spatial correlation was adopted to estimate under-five mortality risk factors. The 2011 Ethiopian Demographic and Health Survey data was used for this study.

Results: The analysis showed that the risk of under-five mortality increases as the family size approaches seven and keeps increasing. With respect to socio-economic factors, the greater the household wealth, the lower the mortality. Moreover, for older mothers, the chance of their child to dying before reaching five is diminishes.

Conclusion: The model enables simultaneous modeling of possible nonlinear effects of covariates, spatial correlation and heterogeneity. Our findings are relevant because the identified risk factors can be used to provide priority areas for intervention activities by the government to combat under-five mortality in Ethiopia.

Show MeSH

Related in: MedlinePlus

Smoothing components for child mortality with A) Number of household members, B) Number of children 5 and under in household, C) Total children ever born, D) Age of respondent at 1st birth, E) Age at first sex, F) Birth order number, G) Age of respondent.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4373119&req=5

Fig2: Smoothing components for child mortality with A) Number of household members, B) Number of children 5 and under in household, C) Total children ever born, D) Age of respondent at 1st birth, E) Age at first sex, F) Birth order number, G) Age of respondent.

Mentions: In addition to categorical effects, there were continuous effects which were handled non-parametrically to the model. These effects are number of household members, number of children at age five and under in a household, age of respondent at 1st birth, age at first sex, total children ever born, age of respondent and birth order number. The result shows that age of respondent at 1st birth, age at first sex, total children ever born, birth order number and age of respondent had non-linear significant effect on child mortality in Ethiopia. The possible nonlinear effects of the continuous covariates after accounting for other variables are presented in Figure 2 together with 95% credible intervals. In the figure, the dashed lines represent the 95% credible intervals. The figures suggest that age of respondents, number of household members, number of children at age five and under in a household, age of respondent at 1st birth, age at first sex, total children ever born and birth order number effects all depart dramatically from linearity. In each panel, the smooth line is the estimated trend from a generalized additive mixed model for the model with spherical Gaussian covariance structure. Figure 2A shows the estimated smooth function of family size and their 95% confidence interval. The y-axis represents the effect of the age term (posterior mean).Figure 2


Structured additive regression models with spatial correlation to estimate under-five mortality risk factors in Ethiopia.

Ayele DG, Zewotir TT, Mwambi HG - BMC Public Health (2015)

Smoothing components for child mortality with A) Number of household members, B) Number of children 5 and under in household, C) Total children ever born, D) Age of respondent at 1st birth, E) Age at first sex, F) Birth order number, G) Age of respondent.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4373119&req=5

Fig2: Smoothing components for child mortality with A) Number of household members, B) Number of children 5 and under in household, C) Total children ever born, D) Age of respondent at 1st birth, E) Age at first sex, F) Birth order number, G) Age of respondent.
Mentions: In addition to categorical effects, there were continuous effects which were handled non-parametrically to the model. These effects are number of household members, number of children at age five and under in a household, age of respondent at 1st birth, age at first sex, total children ever born, age of respondent and birth order number. The result shows that age of respondent at 1st birth, age at first sex, total children ever born, birth order number and age of respondent had non-linear significant effect on child mortality in Ethiopia. The possible nonlinear effects of the continuous covariates after accounting for other variables are presented in Figure 2 together with 95% credible intervals. In the figure, the dashed lines represent the 95% credible intervals. The figures suggest that age of respondents, number of household members, number of children at age five and under in a household, age of respondent at 1st birth, age at first sex, total children ever born and birth order number effects all depart dramatically from linearity. In each panel, the smooth line is the estimated trend from a generalized additive mixed model for the model with spherical Gaussian covariance structure. Figure 2A shows the estimated smooth function of family size and their 95% confidence interval. The y-axis represents the effect of the age term (posterior mean).Figure 2

Bottom Line: With respect to socio-economic factors, the greater the household wealth, the lower the mortality.The model enables simultaneous modeling of possible nonlinear effects of covariates, spatial correlation and heterogeneity.Our findings are relevant because the identified risk factors can be used to provide priority areas for intervention activities by the government to combat under-five mortality in Ethiopia.

View Article: PubMed Central - PubMed

Affiliation: School of Mathematics, Statistics and Computer Science, University of KwaZulu-Natal, Private Bag X01, Pietermaritzburg, Scottsville, 3209, South Africa. ayele@ukzn.ac.za.

ABSTRACT

Background: The risk of a child dying before reaching five years of age is highest in Sub-Saharan African countries. But Child mortality rates have shown substantial decline in Ethiopia. It is important to identify factors affecting under-five mortality.

Methods: A structured additive logistic regression model which accounts the spatial correlation was adopted to estimate under-five mortality risk factors. The 2011 Ethiopian Demographic and Health Survey data was used for this study.

Results: The analysis showed that the risk of under-five mortality increases as the family size approaches seven and keeps increasing. With respect to socio-economic factors, the greater the household wealth, the lower the mortality. Moreover, for older mothers, the chance of their child to dying before reaching five is diminishes.

Conclusion: The model enables simultaneous modeling of possible nonlinear effects of covariates, spatial correlation and heterogeneity. Our findings are relevant because the identified risk factors can be used to provide priority areas for intervention activities by the government to combat under-five mortality in Ethiopia.

Show MeSH
Related in: MedlinePlus