Diabetes mellitus is associated with hyperglycaemia and advanced glycosylation end-products. In the foot, the consequences of chronic or uncontrolled diabetes are micro and macrovascular disease, neuropathy, reduced joint mobility and structural and soft tissue changes that increase the risk of ulcer development and amputation. Diabetes foot assessment currently includes a comprehensive history, neurological and vascular assessments and examination focussed on dermatological and musculoskeletal abnormalities. Whilst these assessments are helpful for predicting ulceration risk, direct identifiers that enable early therapeutic intervention are lacking. The intention of this review was to ascertain if B-mode ultrasound could be clinically applied to identify structural change in the diabetic foot and be utilised as an early predictor of ulceration risk.
Primary databases and grey literature sources were systematically searched. Selection criteria were that the study included a diabetic sample and used B-mode ultrasound to assess soft tissue structures of the foot (plantar skin, plantar fat pad or intrinsic muscles).
Fifteen studies were identified for inclusion (combined diabetic sample of 773). Ultrasound demonstrated reductions in tissue thickness in diabetics compared to non-diabetics under first (p=0.01) and second (p=0.03) metatarsal heads, but not the third (p=0.24). Statistical heterogeneity was high for ultrasound thickness measures under metatarsal heads four/five (I2 65%, 81%) and very high for plantar skin (I2 98%), heel pad (I2 76%) and intrinsic muscles (I2 91%, 81%). Extensor digitorum brevis (EDB) ultrasound measures were significantly thinner in diabetics for all dimension measures compared to healthy controls except one study, which reported no significant differences in EDB thickness.
No direct evidence was found to indicate B-mode ultrasound measures can predict soft tissue changes in the plantar foot in diabetes, although low level studies indicate ultrasound has the potential to identify structural change. Clinical, methodological and statistical heterogeneity limit result applicability. This review highlights the need for robust prospective longitudinal research to examine the predictive validity of this method.