Reassessing high-energy emission correlations in gamma-ray bursts using a large, homogeneous sample of X-ray afterglows

Gamma-ray bursts (GRBs) exhibit diverse X-ray afterglow light-curves, including breaks and plateau phases, whose physical origins remain debated. Previous studies have often relied on small and heterogeneous samples, which limited their statistical power and led to apparently conflicting claims. Most notably, correlations have been suggested between high-energy (E MeV) detection and X-ray afterglow complexity or plateau incidence. We performed a comprehensive large-scale and unbiased statistical analysis of GRB X-ray afterglows to clarify the origins of light-curve complexity and plateau phases, and to reassess their possible connection with high-energy emission. We developed a fully automated, model-independent analysis of the complete -XRT GRB afterglow catalog, comprising more than 1400 events. Our pipeline applies uniform flare removal and segmented power-law fitting, enabling consistent measurements throughout the entire sample and robust statistical inference. Swift We found that both light-curve complexity and plateau incidence are strongly governed by the data-taking starting time of XRT observations (tXRT). When tXRT is ignored, apparent correlations emerge between high-energy emission and X-ray afterglow morphology, but when the data are stratified or controlled for tXRT, these associations vanish. X-ray complexity and plateaus are therefore not directly coupled to high-energy detectability, indicating that early X-ray morphological features are not predictive of high-energy emission. Our results resolve long-standing claims in the literature and highlight a crucial methodological lesson: controlling for tXRT is indispensable in large-sample studies of GRBs. The automated analysis pipeline provides a foundation for reproducible, large-sample inferences and will be essential for exploiting upcoming datasets from missions such as SVOM, Einstein Probe and THESEUS.