Browsing by Author "Er, Şebnem"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Open Access
    Multivariate Extreme Value Theory with an application to climate data in the Western Cape Province
    (2017) Bhagwandin, Lipika; Er, Şebnem
    An understanding of past and current weather conditions can aid in identifying trends and changes that have occurred in weather patterns. This is particularly important as certain weather conditions can have both a positive and a negative impact on various activities in any region. Together with an ever-changing climate it has become markedly noticeable that there is an upward trend in extreme weather conditions. The aim of this study is to evaluate the efficacy of univariate and multivariate extreme value theory models on climate data in the Western Cape province of South Africa. Data collected since 1965 from five weather stations viz. Cape Town International Airport, George Airport, Langebaanweg, Plettenberg Bay and Vredendal was modelled and analysed. In the multivariate analysis, multiple variables are modelled at a single location. Block maxima, threshold excess and point process approaches are used on the weather data, specifically on rainfall, wind speed and temperature maxima. For the block maxima approach, the data is grouped in n-length blocks and the maxima of each block form the dataset to be modelled. The threshold excess and point process approaches use a suitably chosen threshold whereby observations above the threshold are considered as extreme and therefore form the dataset used in the models. Under the threshold excess approach, only observations that exceed the threshold in all components are able to be modelled, whereas exceedances in one and all components simultaneously can be handled by the point process approach. While the probability of experiencing high levels of rainfall, wind speed and temperature individually and jointly are low, a few conclusions were drawn based on the comparison of the performance of the models. It was found that models under the block maxima approach did not perform well in modelling the weather variables at the five stations in both the univariate and multivariate case as many useful observations are discarded. The threshold excess and point process approaches performed better in modelling the weather extremes. Similar results are achieved between these two approaches in the univariate analysis and there is no outright distinction that favours one approach over the other. In terms of the multivariate case, which is restricted to two variables, the point process approach was able to provide estimates with increased accuracy as in many cases there are more extremes in one component individually than in both components. Specifically, the negative logistic and negative bilogistic models suitably capture the dependence structure between maximum wind speed versus maximum rain- fall and maximum wind speed versus maximum temperature at the five weather stations. The results from the point process models showed very weak dependence between wind speed and rainfall maxima as well as between wind speed and temperature maxima which may warrant the inclusion of additional variables into the analysis and even a spatial component which is not included in this study.
  • Thumbnail Image
    Item
    Open Access
    Understanding blood pressure dynamics in the South African population: a latent variables approach to the analysis and comparison of data from multiple surveys
    (2017) Cois, Annibale; Ehrlich, Rodney I; Er, Şebnem
    Background: The 2015 edition of the Global Burden of Diseases Study identified elevated systolic blood pressure─ defined as systolic blood pressure greater than the minimum risk category of 110–115 mm Hg ─ as the largest single contributor to the global burden of disease, responsible for 211.8 million disability adjusted life years lost, up 8.8% in the last decade. Middle‐income countries are currently bearing the highest share of this burden, and, because of the rapid demographic transition towards larger and older populations, the burden is bound to increase rapidly in the coming years, unless age‐specific values of blood pressure are substantially reduced to compensate for the unfavourable demographic changes. Achieving this more favourable blood pressure distribution in populations undergoing rapid changes in their socioeconomic structure requires knowledge of the mechanisms underlying temporal variations of blood pressure and the relationships of such variations with socioeconomic variables.However, evidence on these mechanisms and reliable information on the temporal trends of blood pressure themselves are scant outside high‐income countries. Given the large gain in health that would result in low‐ and middle‐income countries if an optimal blood pressure were to be achieved in large sectors of the population, there is little doubt that temporal trends in the distribution of blood pressure in these populations and their possible determinants are an open and important area for investigation. Objectives: Objectives of the study were: 1. To assess the level of quality and comparability of blood pressure data collected in a series of large‐scale surveys carried out between 1998 and 2015 in South Africa, a middle‐income country undergoing rapid demographic and epidemiological transition; 2. To explore the possibility of applying a series of latent variables techniques to improve the comparability of data from the different sources and to minimise the effect of measurement and representation error on the estimation of cross‐sectional relationships and temporal trends; 3. To estimate changes in the distribution of blood pressure and derived quantities ‐‐‐ such as prevalence of uncontrolled hypertension ‐‐‐ in the South African adult population between 1998 and 2015, taking into account between‐surveys differences and measurement and representation error that could lead to artefactual conclusions; 4. To estimate the extent to which the estimated changes in the blood pressure distribution during the study period could be explained by concurrent changes in the distribution of a series of biological, behavioural and socioeconomic risk factors. Methods: A series of techniques within the general framework of structural equation modelling were applied to jointly analyse the data and estimate the temporal trends and relationships of interest. Results: The average systolic and diastolic blood pressure of South African adult women has progressively decreased since 2003‐2004, reversing the previous rising trend. Among men, the reversal happened only for the systolic blood pressure, while the average diastolic blood pressure continued rising, although at a lower pace than previously.In both genders, this pattern resulted in a reduction of the prevalence of uncontrolled hypertension between 2003‐2004 and 2014‐2015, by 8 percentage points among women and by 4.5 percentage points among men. This consistent and rapid decrease cannot be explained by changes in the age structure of the population, smoking and alcohol consumption habits, distribution of body mass index or urbanization. The diffusion of antihypertensive treatment and, among women, cohort effects and rapidly increasing educational level partly explain the recent trend, but a substantial part of the observed decrease remains unexplained by the factors available in our analyses. Large seasonal variations in both systolic and diastolic blood pressure are present in the South African population, and their magnitude is greater among population strata with low socioeconomic status. From a methodological point of view, there were two further results of this study. First, estimates of blood pressure and related quantities from the eight large‐scale population surveys carried out in South Africa between 1998 and 2015are not directly comparable, because of methodological differences and overall data quality. Second, structural equation modelling (and, within this general framework, multiple group modelling, normal‐censored regression, mixture analysis with skew‐normal distributions and the use of additional parameters and phantom variables) represent a viable and advantageous alternative to current methods of comparative analysis of blood pressure data. Conclusions: Encouraging signs regarding the future development of the burden of diseases related to elevated blood pressure in the South African population emerge from this study. Age‐specific prevalence of uncontrolled hypertension seems to be decreasing, especially among women, and this decrease is accompanied by declining mortality for cardiovascular disease, particularly for stroke, recorded in burden of mortality studies. The reasons of this decrease are largely unexplained and warrant further investigation. However, among the possible drivers analysed in this study, increased accessibility and efficacy of antihypertensive treatment are likely to be playing a role in the observed decrease in blood pressure. The growing obesity epidemic, on the contrary, is likely to be limiting the achievable benefits. Both of these factors can be targeted to maintain and improve the current decline in population values of blood pressure and prevalence of hypertension. The large seasonal variations of blood pressure and their unequal distribution across socioeconomic strata also suggest that interventions to reduce exposure to low temperatures might have public health benefit. From the point of view of the epidemiological investigation, the results of this study suggest that the current methods for the analysis of survey data on blood pressure and the measurement protocols for future data collections should be improved to increase between‐surveys comparabilityand gather more reliable information on temporal changes in BP and gain better understanding of their drivers. Specifically, analytical methods should take explicitly into account known sources of measurement and representation error to reduce their biasing effects, especially when inter‐survey comparisons are involved. Protocols for future studies should routinely include collection of auxiliary information and/or explicit validation of devices and procedures in the specific population.