Identification of pre-clinical cardiovascular imaging phenotypes related to blood pressure in young adults

Thesis

Hypertension prevalence in young adults is rising and is associated with an increased risk of stroke and cardiovascular disease in later life. However, hypertension management guidelines for patients below the age of 40 years remain conservative as their absolute risk of imminent vascular events is low. Additionally, due to the relatively short duration of exposure to hypertension, traditional measures of target organ damage such as left ventricular hypertrophy are often normal. In this thesis, I sought to identify pre-clinical cardiovascular phenotypes associated with blood pressure elevation in young adults, which might be of value for monitoring disease progression in young people.

A cohort of young participants aged 18 to 40 years with a range of blood pressure measures were recruited through clinical studies conducted at the Cardiovascular Clinical Research Facility for multi-dimensional assessment of cardiovascular health. Detailed cardiac structure and function assessment was performed using a range of imaging modalities including resting and stress echocardiography imaging. First, I established that participants with suboptimal blood pressure (≥120/80 mmHg) have novel changes in cardiovascular remodelling identifiable with echocardiography. Specifically, I identified a reduction in left ventricular systolic function during physical exercise, which was associated with altered left atrial pump function at rest. Then, I studied whether it was possible to combine multiple echocardiographic measures using a contrastive trajectory inference machine learning model to describe the progression of cardiovascular remodelling in young adults with hypertension, and place participants on a pseudo-temporal trajectory of disease, with a corresponding score, from health (zero) to disease (one). I demonstrated feasibility of this approach and the practicality of summarising this disease progression as a single score. Finally, I demonstrated that longer duration of treatment was associated with higher score and was consistent with an established modifiable cardiovascular risk score and fitness levels. Therefore, I have demonstrated that there are progressive changes in cardiac structure and function, identifiable with echocardiography, in young adults with suboptimal blood pressure. Further work is now needed to determine whether these measures could be used clinically to better manage blood pressure in younger patients.

Authors: Alsharqi, M

• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: artificial intelligence; young adults; machine learning; cardiac remodelling; hypertension