Endothelial function and vascular health of black South Africans : the significance of plasminogen activator inhibitor-1

Abstract

Background and motivation - Plasminogen activator inhibitor-1 (PAI-1), a serine protease inhibitor known for its classic role in fibrinolysis, has in recent years gained attention for its association with impaired vascular function and blood pressure. On a genetic level, the 4G/5G polymorphism in the promoter region of the PAI-1 (SERPINE1) gene, known to influence PAI-1 levels, has also been associated with blood pressure. It is speculated that microvascular dysfunction not only accompanies, but possibly also precedes, macrovascular dysfunction. Yet limited studies have investigated the association of microvascular function with PAI-1. The observed association of PAI-1 with blood pressure is partly attributed to PAI-1's alternative roles in endothelial dysfunction and vascular remodelling. Whether elevations in PAI-1 precede or result from elevated blood pressure is still an ongoing debate, and prospective studies that may shed light on the sequence of events are limited. The majority of previous studies in this field were performed in older individuals and in patients already presenting with hypertension, with very limited information on the potential role of PAI-1 in the early phases of cardiovascular disease development. Numerous modifiable and non-modifiable factors are known to influence PAI-1 levels. These include health behaviours such as smoking, alcohol consumption and obesity, especially central obesity, and non-modifiable factors such as sex, genetics and Black ethnicity. The extent to which these determinants of PAI-1 influence the associations of PAI-1 with blood pressure and microvascular function is largely unknown. Black South Africans are prone to hypertension development and are known to display unique cardiovascular and haemostatic profiles. The investigation of the associations of PAI-1 (as a potential early marker of vascular dysfunction) with retinal microvascular function and blood pressure across the life course of Black South Africans will aid in elucidating the relationship of PAI-1 with micro- and macrovascular function related to hypertension. Aim - The overarching aim of this study was to investigate the associations of blood pressure and microvascular function with PAI-1 in Black South Africans, and to investigate to what extent these associations are influenced by selected modifiable and non-modifiable factors.

Methodology - Data from two studies were used: The African Prospective study on the Early Detection and Identification of Cardiovascular disease and Hypertension (African-PREDICT), which included Black and White individuals aged 20-30 years, and the South African arm of the Prospective Urban and Rural Epidemiological (PURE) study, which included Black individuals older than 30 years of age. In young, apparently healthy individuals we investigated the associations of plasma PAI-1 activity (PAI-1act) with retinal vasodilatory responses to flicker light provocation. We furthermore investigated the association of PAI-1act with 24-hour blood pressure (24 h BP) and determined the influence of smoking and alcohol use on these associations. In the older cohort of Black individuals we investigated the associations of PAI-1act and the 4G/5G polymorphism with brachial and central blood pressure measures, as well as with hypertension status, in both cross-sectional and prospective analysis (over a 10-year period). Results - In young individuals (aged 20-30 years), maximal retinal venular dilation was independently and inversely associated with PAI-1act (Adj. R2=0.11; β= -0.15; p=0.001) in the total group (n=518). In exploratory subgroup analysis, this association persisted in White women only (Adj. R2=0.07; β= -0.23; p=0.005) and was more robust in non-smokers and individuals of younger age, those with lower blood pressure, those with higher low-density lipoprotein cholesterol and also in individuals with higher levels of inflammation and greater central adiposity (all p<0.05). In the same young cohort (n=1156), 24 h blood pressure measures (systolic and diastolic blood pressure [SBP and DBP] as well as pulse pressure [PP]) were positively associated with PAI-1act in age-adjusted models in most groups (split for sex and ethnicity) (p<0.05). Upon multivariate adjustment and stratification based on self-reported alcohol use and smoking, the 24 h BP-PAI-1act association persisted only in Black men who consumed alcohol (24 h SBP [B=4.22, p<0.001], DBP [B=2.04, p=0.015] and PP [B=2.18, p=0.013]) and smoked (24 h SBP [B=6.10, p<0.001] and PP [B=4.33, p=0.001]). In the older PURE cohort of Black individuals, PAI-1act associated positively with brachial and central blood pressure measures at most time points in cross-sectional analysis. In multivariate analysis, PAI-1act independently and prospectively associated with brachial SBP (r=0.0815) and PP (r=0.0832) in the total group, and with central PP in women (r=0.1125; all p<0.05). Furthermore, both elevated PAI-1act and the 4G/4G (vs the 5G/5G) genotype increased the odds of hypertension development in the total group (1.04 [1.01; 1.08] and

1.82 [1.07; 3.12], respectively). The presence of central obesity either decreased or nullified the associations of PAI-1act with BP and hypertension development. Conclusion - In young apparently healthy adults, PAI-1 was positively associated with 24 h blood pressure and impaired retinal microvascular function. In older individuals, PAI-1 was positively associated with brachial and central blood pressure in both cross-sectional and prospective analysis, and both elevated PAI-1 levels and the 4G/4G genotype increased the odds of developing hypertension. These associations differ in women and men and between ethnicities, and are influenced by modifiable factors such as central obesity, smoking and alcohol consumption. Our data provide evidence that PAI-1 is associated with microvascular