Meta-analysis of longitudinal studies combines effect sizes measured at pre-determined time points. The most common approach involves performing separate univariate meta-analyses at individual time ...points. This simplistic approach ignores dependence between longitudinal effect sizes, which might result in less precise parameter estimates. In this paper, we show how to conduct a meta-analysis of longitudinal effect sizes where we contrast different covariance structures for dependence between effect sizes, both within and between studies. We propose new combinations of covariance structures for the dependence between effect size and utilize a practical example involving meta-analysis of 17 trials comparing postoperative treatments for a type of cancer, where survival is measured at 6, 12, 18 and 24 months post randomization. Although the results from this particular data set show the benefit of accounting for within-study serial correlation between effect sizes, simulations are required to confirm these results.
Abstract Background Childhood stunting is a major indicator of child malnutrition and a focus area of Global Nutrition Targets for 2025 and Sustainable Development Goals. Risk factors for childhood ...stunting are well studied and well known and could be used in a risk prediction model for assessing whether a child is stunted or not. However, the selection of child stunting predictor variables is a critical step in the development and performance of any such prediction model. This paper compares the performance of child stunting diagnostic predictive models based on predictor variables selected using a set of variable selection methods. Methods Firstly, we conducted a subjective review of the literature to identify determinants of child stunting in Sub-Saharan Africa. Secondly, a multivariate logistic regression model of child stunting was fitted using the identified predictors on stunting data among children aged 0–59 months in the Malawi Demographic Health Survey (MDHS 2015–16) data. Thirdly, several reduced multivariable logistic regression models were fitted depending on the predictor variables selected using seven variable selection algorithms, namely backward, forward, stepwise, random forest, Least Absolute Shrinkage and Selection Operator (LASSO), and judgmental. Lastly, for each reduced model, a diagnostic predictive model for the childhood stunting risk score, defined as the child propensity score based on derived coefficients, was calculated for each child. The prediction risk models were assessed using discrimination measures, including area under-receiver operator curve (AUROC), sensitivity and specificity. Results The review identified 68 predictor variables of child stunting, of which 27 were available in the MDHS 2016–16 data. The common risk factors selected by all the variable selection models include household wealth index, age of the child, household size, type of birth (singleton/multiple births), and birth weight. The best cut-off point on the child stunting risk prediction model was 0.37 based on risk factors determined by the judgmental variable selection method. The model’s accuracy was estimated with an AUROC value of 64% (95% CI: 60%-67%) in the test data. For children residing in urban areas, the corresponding AUROC was AUC = 67% (95% CI: 58–76%), as opposed to those in rural areas, AUC = 63% (95% CI: 59–67%). Conclusion The derived child stunting diagnostic prediction model could be useful as a first screening tool to identify children more likely to be stunted. The identified children could then receive necessary nutritional interventions.
Background: South Africa is experiencing an increasing burden of cardiovascular diseases, including coronary heart disease (CHD) and stroke. We aimed to obtain overall national prevalence estimates ...of CHD and stroke in South Africa.Methods: Studies conducted in South Africa were systematically reviewed from PubMed, Scopus and Web of Science from January 1990 to July 2017. Random-effects meta-analyses were conducted on the selected studies to determine the overall prevalence of CHD and stroke.Results: Out of 2 466 studies, only 12 covering 75 140 participants reported the national prevalence of CHD and stroke. All 12 studies estimated the national prevalence of both diseases based on self-reported disease status. The overall national prevalence was 1.29 (95% CI = 0.83; 1.75) and 4.29 (95% CI = 3.13; 5.45) for CHD and stroke, respectively. Only one study reported incidence rates so we did not perform any meta-analysis of incidence rates.Conclusions: There are very few studies on national prevalence of CHD and stroke in South Africa. Well-structured registries for CHD and stroke are required to accurately identify the disease burden and enable adequate resources to be allocated for the implementation of appropriate prevention and management programmes.
Introduction
Consumption of unhealthy foods in children contributes to high levels of childhood obesity globally. In developing countries there is paucity of empirical studies on the association. ...This study employed propensity-score methods to evaluate the effect of unhealthy foods on overweight among children in Malawi using observational data.
Methods
Data on 4625 children aged 6 to 59 months from the 2015-16 Malawi Demographic and Health Survey (MDHS) were analyzed. A multivariable logistic regression model of unhealthy foods (yes or no) on purported confounders of childhood overweight was used to obtain a child’s unhealthy food propensity score. The propensity scores were then used to form matched sets of healthy and unhealthy fed children. The association between unhealthy foods and childhood overweight was assessed using the conditional logistic regression model.
Results
The prevalence of overweight (body mass index (BMI) z-score > 2 standard deviations) was estimated at 4.5% (3.8%, 5.3%). The proportion of children who consumed unhealthy foods was estimated at 14.6% (95% CI: 13.1%, 16.2%). Our propensity score matching achieved a balance in the distribution of the confounders between children in the healthy and unhealthy food groups. Children fed unhealthy foods were significantly more likely to be overweight than those fed healthy foods (OR = 2.5, 95% CI: (1.2, 5.2)).
Conclusion
The findings suggest the adverse effects of unhealthy foods on childhood overweight in Malawi. Thus, efforts to reduce unhealthy food consumption among children should be implemented and supported to address the problem of childhood overweight in Malawi and the sub-Saharan African region.
Although under-five mortality (U5M) rates have declined worldwide, many countries in sub-Saharan Africa still have much higher rates. Detection of subnational areas with unusually higher U5M rates ...could support targeted high impact child health interventions. We propose a novel group outlier detection statistic for identifying areas with extreme U5M rates under a multivariate survival data model. The performance of the proposed statistic was evaluated through a simulation study. We applied the proposed method to an analysis of child survival data in Malawi to identify sub-districts with unusually higher or lower U5M rates. The simulation study showed that the proposed outlier statistic can detect unusual high or low mortality groups with a high accuracy of at least 90%, for datasets with at least 50 clusters of size 80 or more. In the application, at most 7 U5M outlier sub-districts were identified, based on the best fitting model as measured by the Akaike information criterion (AIC).
Statistical techniques for detecting influential data are well developed and commonly used in linear regression, and to some extent in linear mixed-effects models. However, even though the ...application of multivariate survival models is widely undertaken, the development of diagnostic tools for the models has received less attention. In this article, we extend the martingale-based residuals and leverage commonly used in univariate survival regression to derive influence statistics for the multivariate survival model. The performance of the proposed statistic is evaluated by simulation studies. The statistic is illustrated with an analysis of child clustered survival data to identify influential clusters of observations and their effects on the estimate of fixed-effect coefficients.
Abstract Background Several studies have shown that maternal HIV infection is associated with adverse pregnancy outcomes such as low birth weight and perinatal mortality. However, the association is ...conflicted with the effect of antiretroviral therapy (ART) on the pregnancy outcomes and it remains unexamined. If the association is confirmed then it would guide policy makers towards more effective prevention of mother to child HIV transmission interventions. Using methods for matching possible confounders, the objectives of the study were to assess the effect of maternal HIV infection on birth weight and perinatal mortality and to investigate the effect of ART on these two pregnancy outcomes in HIV-infected women. Methods Data on 4111 and 4759 children, born within five years of the 2010 and 2015-16 Malawi Demographic and Health Surveys (MDHS) respectively, whose mothers had an HIV test result, were analysed. A best balancing method was chosen from a set of covariate balance methods namely, the 1:1 nearest neighbour (NN) matching, matching on the propensity score (PS) and inverse weighting on the PS. HIV and ART data were only available in the MDHS 2010, permitting an assessment of the moderating effect of ART on the association between maternal HIV infection and birth weight and perinatal mortality. Results The overall average birth weight was 3227.9g (95% CI: 3206.4, 3249.5) in 2010 and 3226.4g (95%: 3205.6, 3247.2) in 2015-16 and perinatal mortality was 3.8% (95%: 3.2, 4.3) in 2010 and 3.5% (95%: 2.8, 3.8) in 2015-16. The prevalence of HIV among the mothers was 11.1% (95%: 10.1, 12.0) and 9.2% (95% CI: 8.4, 10.1) in 2010 and 2015-16, respectively. In 2010, maternal HIV infection was negatively associated with birth weight (mean= -25.3g, 95% CI:(-95.5, -7.4)) and in 2015-16 it was positively associated with birth weight (mean= 116.3g, 95% CI:(27.8, 204.7)). Perinatal mortality was higher in infants of HIV-infected mothers compared to infants of HIV-uninfected mothers (OR = 1.5, 95% CI:(1.1 - 3.1)) in 2010, while there was no difference in the rate in 2015-16 (OR = 1.0, 95% CI:(0.4, 1.6)). ART was not associated with birth weight, however, it was associated with perinatal mortality (OR=3.9, 95% CI:(1.1, 14.8)). Conclusion The study has found that maternal HIV infection had an adverse effect on birth weight and perinatal mortality in 2010. Birth weight was not dependent on ART uptake but perinatal mortality was higher among infants of HIV-infected mothers who were not on ART. The higher birth weight among HIV-infected mothers and similarity in perinatal mortality with HIV-uninfected mothers in 2015-16 may be indicative of successes of interventions within the PMTCT program in Malawi.
To profile the prevalence of the three body mass index (BMI) categories by sociodemographic characteristics, and to calculate the percentage transitioning (or not) from one BMI category to another, ...to inform South African health policy for the control of obesity and noncommunicable diseases.
We used data from the National Income Dynamics Study, including sociodemographic characteristics and BMI measurements collected in 2008, 2010, 2012, 2014 and 2017. For each data collection wave and each population group, we calculated mean BMI and prevalence by category. We also calculated the percentage making an upwards transition (e.g. from overweight to obese), a downwards transition or remaining within a particular category. We used a multinomial logistic regression model to estimate transition likelihood.
Between 2008 and 2017, mean BMI increased by 2.3 kg/m
. We calculated an increased prevalence of obesity from 19.7% (3686/18 679) to 23.6% (3412/14 463), with the largest increases in prevalence for those aged 19-24 years and those with at least high school education. The percentages of upwards transitions to overweight or obese categories increased sharply between the ages of 19 and 50 years. Once overweight or obese, the likelihood of transitioning to a normal BMI is low, particularly for women, those of higher age groups, and those with a higher income and a higher level of education.
In the development of national strategies to control obesity and noncommunicable diseases, our results will allow limited public health resources to be focused on the relevant population groups.
PDF
