A prospective study of risk factors for diabetic foot ulcer. The Seattle Diabetic Foot Study.

OBJECTIVE

Little prospective research exists on risk factors for diabetic foot ulcer that considers the independent effects of multiple potential etiologic agents. We prospectively studied the effects of diabetes characteristics, foot deformity, behavioral factors, and neurovascular function on foot ulcer risk among 749 diabetic veterans with 1,483 lower limbs.

RESEARCH DESIGN AND

METHODS

Eligible subjects included all diabetic enrollees of a general internal medicine clinic without foot ulcer, of whom 83% agreed to participate. Baseline assessment included history and lower-limb physical examination, tests for sensory and autonomic neuropathy, and measurements of macro- and microvascular perfusion in the foot. Subjects were followed for the occurrence of a full thickness skin defect on the foot that took > 14 days to heal, with a mean follow-up of 3.7 years.

RESULTS

Using stepwise Cox regression analysis, the following factors were independently related to foot ulcer risk: foot insensitivity to the 5.07 monofilament (relative risk [95% CI]) 2.2 (1.5-3.1), past history of amputation 2.8 (1.8-4.3) or foot ulcer 1.6 (1.2-2.3), insulin use 1.6 (1.1-2.2), Charcot deformity 3.5 (1.2-9.9), 15 mmHg higher dorsal foot transcutaneous PO2 0.8 (0.7-0.9), 20 kg higher body weight 1.2 (1.1-1.4), 0.3 higher ankle-arm index 0.8 (0.7-1.0), poor vision 1.9 (1.4-2.6), and 13 mmHg orthostatic blood pressure fall 1.2 (1.1-1.5). Higher ulcer risk was associated with hammer/claw toe deformity and history of laser photocoagulation in certain subgroups. Unrelated to foot ulcer risk in multivariate models were diabetes duration and type, race, smoking status, diabetes education, joint mobility, hallux blood pressure, and other foot deformities.

CONCLUSIONS

Certain foot deformities, reduced skin oxygenation and foot perfusion, poor vision, greater body mass, and both sensory and autonomic neuropathy independently influence foot ulcer risk, thereby providing support for a multifactorial etiology for diabetic foot ulceration.