Wearable technologies enable continuous, non-invasive monitoring of cardiovascular health and hold great promise for tracking dynamic physiological responses during activities such as exercise. Photoplethysmography (PPG) is a widely use sensing modality due to its simplicity and cost-effectiveness, but its accuracy during movement remains a challenge. This study aims to investigate the effect of motion-induced hemodynamic changes on in-ear PPG pulse wave morphology during heart-paced walking. We hypothesize that motion artifacts partly originate from physiological effects induced by body motion, such as blood inertia and wave reflections. We distinguish these physiological effects from motion artifacts, by tightly synchronizing walking to the heart pace and performing a rigorous selection of beat-wise pulse waveforms. The 12 healthy participants (6 female, 28±2 years) took part in the study walked on a treadmill at a comfortable speed, guided by auditory signals to synchronize their steps to either systole (R-wave) or diastole (45% RR interval) of the cardiac cycle. In-ear PPG, electrocardiogram and chest vertical acceleration were simultaneously acquired during walking and rest. PPG pulse wave amplitude and morphology varied significantly during heart-paced walking. Diastolic stepping resulted in a consistent morphology downward deflection after half of the cardiac cycle, while systolic stepping results in a significant 0.06V increase in peak-to-peak amplitude compared to diastolic stepping, with large effect size (Cohen's d=1.08). Heart rate does not affect their interaction. These findings highlight the potential of in-ear PPG to capture physiological changes during dynamic conditions. However, further studies are needed on isolating and extracting motion-induced hemodynamic changes from the cardiac signal in less controlled and non-synchronized condition.