Exploring the strength of relation of cardiovascular health with future cardiovascular events using Life Simple Seven and Life Essential Eight approach Master of Science degree in Epidemiology (Medicine faculty-Utrecht University): Major research project Name student: Aleksandra Danailova (2024) Name tutor: prof. Dr. Michiel L. Bots, MD, PhD, University Medical Center Utrecht, Utrecht University Second reviewer: Dr. Ewoud Schuit, PhD, University Medical Center Utrecht

Abstract

CVD severely impacts the quality of Life and mortality (McKearnan, Wolfson, Vock, Vazquez-Benitez & O'Connor, 2018). This calls for urgent preventive measures, so diverse models are proposed to tackle this problem. AHA developed criteria for ideal health dubbed the Life Simple Seven (LS7) and Life Essential Eight (LE8) to quantify population health based on health behaviours and factors (Lloyd-Jones et al., 2022; Lloyd-Jones et al., 2022a). When these factors are optimized, they are associated with better CVD-free survival, total longevity, and improved quality of Life. The project aims to assess to which extent the LS7 or LE8 model better relates to the risk of developing CVD. Using information from the individual participant data of various international population cohorts in the USE-IMT initiative , we quantified cardiovascular health into ideal, intermediate/moderate and poor health based on LS7 and LE8 approaches. Cox semi-parametric models were used to estimate the HRr-hazard rates for CVD. The HR-hazard ratios of CVD for the ideal versus the poor category and intermediate versus poor were then calculated separately based on a couple of inputs for LS7 and LE8. The two main models were compared using the net reclassification improvement (NRI) method to explore which model best predicts the risk. We used six (6) components (health behaviours and factors) of the CVH approach using data from 31,549 participants from 16 cohorts (mean age 57.80 yrs, 53% women). LS7's hazard ratio of ideal versus poor health for developing a cardiovascular event is 0.25 with a CI (0.207-0.306), and for the intermediate category versus poor is 0.437 with a CI (0.360-0.530). The corresponding hazard ratio based on LE8 is 0.237 with a CI of (0.209-0.269) and o.468 with a CI of (0.416-0.527) respectively. Individually, the metrics differ in the magnitude and some in the direction of effects. Based on the analysis performed with NRI, the net reclassification improvement index (NRI) is 0.098. In conclusion: the new LE8 model is slightly more accurate than the LS7, meaning LE8 classified subjects approximately 10% more accurately/correctly than the old model, FS7. So, LE8 is the preferred model for use. Further research could build on the findings by exploring the potential applications of the LE8 model in clinical settings.