Enhancement of precision and reliability in simulation models requires the exploration of advanced sampling techniques and parameter optimization strategies. This study integrates Latin Hypercube Sampling (LHS) with parameter adjustment methodology to explore the variability of simulation outputs, focusing on an Agent-Based Model (ABM) for atherosclerotic plaque progression sensitive to parameter variations. The research emphasizes improving model's generalization capabilities by systematically varying parameters within defined ranges, including constant and dynamic parameters, to evaluate their impact on the simulation results. LHS was used to generate a comprehensive parameter space, consisting of 7 distinct parameters, with two set as constants and the remaining five subject to variation within specified ranges. This approach facilitated the generation of diverse, yet evenly distributed, parameter sets to feed into the ABM simulations. The application of LHS in conjunction with the parameter adjustment strategy revealed significant insights into the model's sensitivity to various parameters. The findings indicate a pronounced impact of certain parameters on the simulation outcomes, specifically the parameter driving SMC proliferation in intima, increased probability of lipid infiltration and increased ECM degradation highlighting their potential as critical levers impacting model's behavior, suggesting areas for further investigation and optimization. By systematically exploring the parameter space and identifying key parameters with pronounced impacts on model outcomes, this research lays the groundwork for further refinement of simulation practices. Future work will delve deeper into the mechanisms underlying the observed sensitivities and extend the approach to encompass a wider array of models and scenarios.