The Interplay between Fasting Glucose, Echocardiography, and Biomarkers: Pathophysiological Considerations and...

Traditional cardiovascular risk stratification tools that employ clinical risk factors are limited by their modest discriminative abilities. As such, robust cardiovascular risk assessment, including our understanding of the complex interplay between risk factors, in the primary preventive setting, remains incomplete. Phenotypical heterogeneity may be even greater among subjects with hyperglycemic conditions, i.e., prediabetes and diabetes, which is worrisome, given the dramatic global rise in mean fasting glucose levels, and the strong association with adverse cardiovascular outcomes. The unmet need for refinement or restratification of risk based on these conventional prediction models is only emphasized by our entrance into the era of precision medicine. Potential tools for closing these gaps and increasing our understanding of the pathways from risk factors through subclinical changes to manifest disease include echocardiography and circulating biomarkers.

1) To examine whether greater fasting plasma glucose (FPG) levels were associated with left ventricular mass (LVM), geometric pattern, diastolic function, and concentrations of N-terminal prohormone of brain natriuretic peptide (NT-proBNP) and high-sensitivity cardiac troponin T (hs-TnT) in apparently healthy, elderly subjects with a preserved LV ejection fraction ± 50%; 2) To examine whether FPG levels modified the prognostic role of abnormal LVM, geometric pattern, diastolic dysfunction, NT-proBNP, and hs-TnT, in predicting cardiovascular morbidity and mortality; 3) To define the incremental prognostic value of NT-proBNP and hs-TnT for predicting incident cardiovascular outcomes, beyond traditional risk factors, glycemic status, and subclinical echocardiographic abnormalities; 4) To explore the associations of NT-proBNP and hs-TnT with key echocardiographic measures of LV structure and function, including the effects of FPG levels.

The thesis was based on a series of cross-sectional and prospective observational studies. The study population was derived from the echocardiography subsample (n=1,792) of the Malmö Preventive Project Re-Examination Study (MPP-RES) (2002-2006, n=18,238), a population-based screening program that included inhabitants from Malmö, Sweden, who belonged to prespecified birth cohorts between 1921-1949. Subjects, who underwent echocardiography, were randomly chosen from the three categories defined by base-line FPG, i.e., normal fasting glucose, impaired fasting glucose, and diabetes, including use of anti-diabetic medication. Blood samples for cardiovascular biomarker assessments were drawn at the time of echocardiography and kept frozen until analysis. Outcome data were obtained through national and local registries. The original echocardiography subsample was stratified into patients and apparently healthy subjects, the latter being the focus of this thesis.

1) Subjects with diabetes had a greater prevalence of concentric LV hypertrophy (LVH), grade 2 or 3 diastolic dysfunction, and higher hs-TnT concentrations. Subjects with impaired fasting glucose had the lowest NT-proBNP concentrations. LVMI was primarily associated with diastolic function in subjects with hyperglycemia; 2) LV diastolic dysfunction was associated with an increased risk of incident cardiovascular events, but did not provide discriminative im-provement. Concentric LVH and diastolic dysfunction were more strongly associated with adverse prognosis in subjects with hyper-glycemia. High concentrations of NT-proBNP and hs-TnT predicted incident cardiovascular events, with no effect modification by FPG; 3) NT-proBNP, but not hs-TnT, provided discriminative improvement beyond traditional risk factors, FPG, and LVH and/or diastolic dysfunction; 4) NT-proBNP and hs-TnT were associated with several echocardiographic parameters, but effect sizes were generally modest. Associations between biomarkers and echocardiographic measures were affected by hyperglycemia.

FPG influenced the interplay between subclinical echocardiographic abnormalities, circulating biomarkers, and cardiovascular outcomes at multiple stages, in this cohort of apparently healthy, elderly subjects. Newly diagnosed diabetes, but not impaired fasting glucose, was associated with adverse subclinical changes. The associations between structural echocardiographic abnormalities and biomarker concentrations were stronger in subjects with hyperglycemia. NT-proBNP, but not echocardiographic measures or hs-TnT, provided discriminative improvement on top of traditional cardiovascular risk factors. FPG further modified the prognosis re-lated to echocardiographic alterations, but not that predicted by biomarkers. Therefore, FPG should be considered when assessing markers of subclinical cardiovascular alterations.