<p>Age-related odor in elderly solo housing may discourage visitor return, yet causal evidence remains absent. A two-stage design was employed, combining field measurement in 35 elderly solo dwellings in Fuyang, China (TVOC concentration, entrance buffer conditions, monthly visitor frequency) with a split-plot experiment exposing 75 participants to standardized (E)-2-nonenal across three design conditions (baseline/ventilation/ventilation-plus-buffer), measuring visitor behavioral intention (VBIS), affect, subjective comfort, and psychophysiological indices (EDA, HRV). Odor level significantly reduced VBIS scores, with effects concentrated in the no-odor to low-odor transition, suggesting a front-loaded, decelerating pattern that was most pronounced under the ventilated and combined design conditions. Affect partially mediated the odor-behavior pathway. Under high odor exposure, the combined design package, consisting of ventilation, an entrance buffer, and enclosed storage, sustained higher behavioral intention than ventilation alone. This finding supports a package-level buffering effect, although the design cannot isolate the independent contribution of any single component. This behavioral buffering was not matched physiologically: a pre-specified equivalence test did not confirm that subjective comfort returned to baseline, and EDA and HRV remained significantly elevated, so questionnaire-only assessment may underestimate the residual physiological burden. The controlled experiment constituted the primary causal test, and the field measurement of 35 dwellings provided a directional real-world reference; in that field stage an exploratory analysis found bedroom TVOC negatively and entrance buffer presence positively associated with monthly visitor frequency, converging in direction with the experiment. These results carry direct implications for olfactory environment management in elderly housing.</p>