Diabetic foot, one of the most common and debilitating manifestations of type 2 diabetes mellitus (T2DM), is the leading cause of worldwide non-traumatic lower extremity amputations. Diabetics who are at risk of ulceration are currently mainly identified by a thorough clinical examination of the feet, which typically does not show clear symptoms during the early stages of disease progression. In this study, we used a non-linear dynamics tool, gait entropy (GaitEN), in addition to traditional linear gait analysis methods, to investigate gait alterations amongst diabetic patients with combinations of three types of T2DM related complications: retinopathy, diabetic peripheral neuropathy (DPN) and nephropathy. Peak plantar pressure (PPP) was not significantly different in the group with DPN as compared to the control group (diabetics with no complications, CONT) in the forefoot region (DPN: mean ± SD: 396 ± 69.4 kPa, CONT: 409 ± 68.9 kPa), although it was significantly lower in the heel region (DPN: mean ± SD: 285 ± 43.1.4 kPa, CONT: 295 ± 61.8 kPa). On the other hand, gait entropy was significantly lower for the DPN compared to CONT group (DPN: 0.95 ± 0.34, CONT: 1.03 ± 0.28, p < 0.05). The significant low entropy was maintained when neuropathy was combined with either retinopathy or nephropathy. For the group with all three complications (ALL-C), the entropy was higher than CONT (ALL-C: 1.07 ± 0.26). This may indicate an intrinsic sensorimotor feedback mechanism for the DPN patients to regulate their gait. However, this feedback gets weaker as patients develop multiple complications. Further analysis with longer walking time and different speeds is needed to verify the entropy results.