Brain function is shaped by the local and global connections between its dynamical units and biological parameters. As we age, the anatomical topology undergoes significant deterioration (e.g., long-range white matter fiber loss), affecting overall brain function. Despite the structural loss, existing studies have pinpointed that normative patterns of functional integrity, defined as the compensatory mechanism of the aging brain, remain intact across the lifespan. However, the crucial components in guiding the adaptive mechanism by which the brain readjusts its bio-logical parameters to maintain optimal compensatory function with age still needs to be uncovered. Here, we provide a parsimonious mechanism, which, together with the data-driven whole-brain generative model, establishes an individualized structure-function link with aging and uncovers the role of the subcommunity in driving the neurocompensation process. We use two neuroimaging datasets of healthy human cohorts with large sample sizes to systematically investigate which of the brain sub-graphs (connected via short or long-range white matter tracts) drives the compensatory mechanisms and modulates intrinsic global scaling parameters, such as interaction strength and conduction delay, in preserving functional integrity. The functional integrity is evaluated under the hypothesis of preserved metastability, measured from individual fMRI BOLD signals. Our findings uncover that the sub-graph connected via short-range tracts mainly modulates global coupling strength to compensate for structural loss. In contrast, long-range connections contribute to the conduction delay, which may play a complementary role in neurocompensation. For the first time, these findings shed light on the underlying neural mechanisms of age-related compensatory mechanisms and provide a mechanistic explanation for the importance of short-range connections in the face of the loss of long-range connections during aging using BOLD fMRI data. This crucial insight could open an avenue to understanding the role of subgraphs for targeted interventions to address aging-associated neurodegenerative diseases where long-range connections are significantly deteriorated.