Fibrinogen concentration and plasma clot properties as predictors of cardiovascular disease mortality in Black South Africans

Abstract

Introduction and aim: Cardiovascular disease (CVD) is one of the main causes of mortality, globally and in South Africa. Recent evidence obtained from prospective studies point to a potential causal role for clot structure in the development of CVD, although less is known regarding CVD mortality. The limited available studies on mortality explored different populations/ethnic groups and were subject to a wide array of inclusion criteria, whilst simultaneously being differentiated by sex, restricting firm conclusions. Prospective evidence, particularly on a population level, is still lacking. We aimed to determine the prospective association between total and ’ (a splice variant of) fibrinogen concentration and plasma clot properties and incident CVD mortality over 13 years in a Black South African cohort. Methods: Data was collected from 2 010 apparently healthy Black South Africans who took part in the South African, North West Province arm of the International Prospective Urban and Rural Epidemiology Study (PURE-SA-NW). Data was collected at three time points; 2005, 2010 and 2015, of which only the baseline data is included in this mini-dissertation. Mortality status and cause of death data was obtained from Statistics South Africa and was available for all participants up until December 2018. The cause of death was documented on death certificates by a medically qualified general practitioner according to the 10th revision of the International Classification of Disease codes (ICD-10 codes). Plasma fibrinogen was measured using the Clauss method, and ’ fibrinogen with an enzyme-linked immunosorbent assay. Markers of plasma clot formation, structure and lysis were measured using the turbidity-based global fibrinolytic assay. Ordinary Cox models, adjusted for known CVD risk markers were used to determine the relationship of total and ’ fibrinogen and plasma clot properties at baseline with incident CVD mortality. CVD mortality was defined in two ways. First CVD mortality was classified when the underlying cause of death was indicated as CVD (ICD10 Codes I00 – I99), this is termed CVDUnderlying (n=106 cases). A second approach was to classify CVD mortality when either underlying cause of death and/or the immediate cause of death was indicated as CVD (ICD10 Codes I00 – I99), this is termed CVDTotal (n=154). Association with all-cause mortality, although not the main aim of this mini-dissertation, is also reported.

Results: Originally, 2 010 participants were included in the PURE population. However, participants were removed from the dataset if they did not have complete data for the exposure variables (coagulation markers). After cleaning the dataset and excluding the participants with missing data the final sample size was n=1 645. Certain infectious and parasitic diseases (ICD 10 codes A00 to B99) (n=144) and Diseases of the circulatory system (CD 10 codes I00 to I99) (n=106) were the two main groups of underlying cause of death in this cohort. The three most prominent underlying CVD causes of death according to the ICD-10 code broad groups were hypertensive diseases (n=37), cerebrovascular diseases (n=31) and other forms of heart disease (n=29). The CVD mortality group consisted of proportionally more individuals from urban areas, had a higher percentage of smokers and alcohol consumers and lower physical activity than the survivors at baseline. They also had a higher prevalence of hypertension, and higher lipoprotein(a), glycated haemoglobin, inflammatory markers, gamma-glutamyl transferase, creatinine and glomerular filtration rate compared to the survivors. Of the coagulation markers, only total fibrinogen and clot density (maximum absorbance) had increased risk for CVDUnderlying and CVDTotal mortality, in a minimally adjusted model (age and gender), while significance was lost in the fully adjusted model. Conclusion: Infectious and parasitic diseases followed by diseases of the circulatory system were the main causes of death in the PURE-SA-NW cohort, while hypertensive, cerebrovascular and heart diseases were the main CVD causes of death. Total fibrinogen and clot density contributed to an increased risk of incident CVD mortality before, but not after adjustment for known CVD risk markers. These findings suggest that fibrinogen concentration and plasma clot properties did not confer additional, independent CVD mortality risk in this Black South African cohort. Although evidence from highly controlled case-control studies suggest a causal relationship for clot properties in CVD progression and mortality, the contribution of fibrinogen and plasma clot properties against the background of other CVD risk factors, such as urban living conditions, on epidemiological level, appears to be less prominent.