Studies of mortality risk predictors in hypertensive patients

Panniyammakal, Jeemon (2013) Studies of mortality risk predictors in hypertensive patients. PhD thesis, University of Glasgow.

Abstract

Hypertension is a leading cause of death and disability worldwide. Despite effective treatment regimens the mortality among hypertensive subjects are much higher than that of normal individuals. Several studies have been carried out to identify prognostic factors that have predictive value for mortality in the general population. New biomarkers that are readily available and cost-effective are important in risk stratification and management of hypertension. While important prognostic information can be learned from routine blood tests that are often conducted in hypertension clinics, the usefulness of these markers in predicting survival are not studied in detail. The thesis aims to explain the relationship between such inexpensive and commonly available markers and survival in a hypertensive population.
The thesis is divided into five main results chapters (chapters 3 to 7) based on studies conducted to assess the independent role of blood pressure variability (BPV), haematocrit, serum phosphate, serum electrolytes and indices of liver dysfunction or injury in predicting mortality in hypertensive patients. The study settings (Glasgow Blood Pressure Clinic) provided an opportunity to examine these relationships in a treated hypertensive cohort of more than 15,000, predominantly white population, from the West of Scotland. The hypertension clinic database was linked with the electronic records of General Register Office for Scotland. This electronic linking allowed extraction of primary cause of death data (if patients died during the course of follow-up) according to the International Classification of Diseases, 10th Revision, Version for 2007 (ICD-10), codes. The type of mortality was ascertained (namely; ischaemic heart disease, stroke, cardiovascular, non-cardiovascular and all-cause) from the ICD-10 codes. The independent relationships between predictor variables of interests and mortality were estimated after employing appropriate survival models. The main study findings are summarised below.
Blood pressure variability and mortality: Long term average BPV is an independent predictor of mortality. Longitudinal changes in BPV also predict mortality independent of underlying mean BP. While sustained high variability increases mortality, sustained low variability decreases mortality in this hypertensive cohort. The findings indicate that BPV is likely a fundamental physiologic trait and it is a marker of early mortality. Visit-to-visit BPV is an important prognostic indicator of long-term mortality, and physicians should be made aware that long term clinic BPV should not be disregarded as random fluctuation between visits.
Haematocrit and mortality: Haematocrit (Hct) is the proportion of blood volume occupied by red blood cells. It is associated with follow-up BP and is an independent predictor of mortality in the hypertensive population. There are distinct differences both in terms of the strength and magnitude of the association of Hct and mortality between men and women that have not previously been known. While Hct is associated with CV mortality in men ('U' shaped, non-linear), it is more closely associated with non-CV mortality in women ('U' shaped, non-linear). In the assessment and management of newly diagnosed hypertensive patients, Hct levels should be taken into consideration as an important risk predictor.
Serum phosphate and mortality: Inorganic phosphate is an important mineral that is directly linked to energy metabolism, bone mineralisation, signal transduction, storage and translation of genetic information and maintenance of lipid membrane structure. A positive linear association between serum phosphate and mortality is reported in the present study. Deprivation status, serum calcium and serum alkaline phosphatase levels do not attenuate the mortality risk associated with serum phosphate in men and women. While serum phosphate is associated with CV mortality in men, it is more closely associated with non-CV mortality in women.
Serum electrolytes and mortality: Electrolytes, especially sodium, chloride, potassium and bicarbonates, play a vital role in maintaining homeostasis within the human body. While the relationship with all-cause mortality is non-linear across the entire range of serum chloride, there is a linear increase in mortality with decrease in serum chloride level below 100 mEq/L. The relationship between serum chloride and mortality is independent of serum sodium and bicarbonate levels. While serum potassium shows a non-linear "U" shaped relationship with mortality, serum bicarbonate shows a positive linear association.
Indices of liver dysfunction or liver injury and mortality: Serum albumin, bilirubin, alanine transaminase (ALT), aspartate transaminase (AST), gamma-glutamyl transferase (GGT) and alakaline phosphatase (ALP) are widely used markers of liver function or injury to liver cells. These markers of liver function or injury to liver cells independently predict mortality outcomes in the hypertensive population. While there is a linear association of both GGT and ALP with mortality outcomes, it is a more complex, non-linear and ‘U’ shaped association for AST. Both ALT and bilirubin show inverse linear association with mortality. Age and body mass index significantly influence the relationship between ALT and mortality.
Strengths and limitations: The strengths of the studies conducted as part of this thesis include; a large cohort of nearly 15,000 hypertensive adults, a real life clinical setting, 35 years of follow-up with median survival time of 32 years, the ability to link predictor variables with differing causes of mortality outcomes, and adjustment for several potential confounding factors. Exclusion of individuals without predictor variables assessed at baseline and the bias introduced by the missing covariates in the adjusted Cox-proportional hazard models are the major weaknesses.
Future recommendations: Although the above mentioned inexpensive markers predict mortality in hypertensive population, the mechanism involved in their association with mortality is not clear. Future studies are required to explain the missing links. Usefulness of inclusion of these markers in predicting mortality should be tested in an independent population.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Additional Information: Chapter 4, Haematocrit and Mortality is published in "Hypertension". Paul L, Jeemon P, Hewitt J, McCallum L, Higgins P, Walters M, McClure J, Dawson J, Meredith P, Jones GC, Muir S, Dominiczak AF, Lowe G, McInnes GT, Padmanabhan S. Hematocrit predicts long-term mortality in a nonlinear and sex-specific manner in hypertensive adults. Hypertension. 2012 Sep;60(3):631-8.
• Keywords: Hypertension, Biomarkers, Mortality, Cardiovascular Disease
• Subjects: R Medicine > R Medicine (General); R Medicine > RA Public aspects of medicine
• Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Cardiovascular & Metabolic Health
• Supervisor's Name: Padmanabhan, Dr. Sandosh
• Date of Award: 2013
• Depositing User: Mr Jeemon Panniyammakal
• Unique ID: glathesis:2013-4031
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 14 Mar 2013 15:35
• Last Modified: 14 Mar 2013 15:37
• URI: https://theses.gla.ac.uk/id/eprint/4031