Quantifying the directional interactions between ecological quality and anthropogenic intensity is crucial for sustainable development. However, the nonlinear response characteristics and threshold effects of human-nature systems remain inadequately explored. This study conducted a spatiotemporal coupling assessment of human--nature relationships in the Henan section of the Yellow River Basin from 2000 to 2020 within the "Pattern-Interaction-Mechanism-Threshold" framework. Ecological quality was represented by the Remote Sensing Ecological Index (RSEI) constructed from MODIS data, while human activity intensity was quantified using Human Footprint (HF) data. The coupling coordination degree model was adopted to evaluate human-nature interactions, and the Geodetector model was employed to reveal the key driving factors. Furthermore, the constraint line and the piecewise linear regression methods were applied to identify ecological response thresholds. The results indicated that the RSEI decreased from 2000 to 2020, reaching its lowest point in 2015. Despite a significant negative spatial correlation between the RSEI and HF, both increased concurrently during 2000-2010 and 2015-2020. Over 58 % of the basin exhibited imbalanced coupling, with human-dominated areas expanding. Distance to major roads and temperature were the key drivers influencing the coupling level, with the explanatory power consistently above 0.25. The ecological response threshold followed a "rise-decline-rebound" trend and is projected to reach 32.7 by 2030, exceeding the 2010 peak. Special attention should be directed to HF Level III regions (30 <= HF < 40) in the central and northern parts to prevent degradation risks from sustained disturbances. These findings provide a scientific basis for the ecological management of sensitive regions.