Design: regional registry-based cohort study using linked routine healthcare data, 2015–2024
Population: 140 adults with diabetes and a first recorded Charcot neuro-osteoarthropathy diagnosis, from 127,513 adults with diabetes
Primary outcome: emergency hospitalisation, lower-extremity amputation, all-cause mortality and amputation-free survival
Effect: 87.9% emergency admission, 31.4% amputation, 37.9% death over median 4.6 years; 5-year survival 67%, amputation-free survival 44%

A registry study linked routine NHS data across Greater Glasgow and Clyde from 2015 to 2024 and identified 140 adults with a first recorded Charcot neuro-osteoarthropathy diagnosis among 127,513 adults with diabetes. Mean age was 59 years and 67% were men. Over a median 4.6 years, 123 (87.9%) had an emergency admission, 44 (31.4%) had a lower-extremity amputation and 53 (37.9%) died. Five-year survival was 67%; five-year amputation-free survival was 44%, with a median of 4.04 years.

The risk was not spread evenly. Advanced chronic kidney disease (stages 4 and 5) was associated with worse outcomes across the board. High or active foot-risk status was independently associated with amputation (subdistribution hazard ratio 8.52, 95% CI 2.05 to 35.45) and with shorter amputation-free survival (HR 2.09, 95% CI 1.15 to 3.80). Older age was independently associated with mortality (HR 3.04, 95% CI 1.50 to 6.16).

Charcot is often managed as an orthopaedic problem about offloading a deformity. These numbers argue for reading it as a marker of systemic vulnerability — the renal function, the cardiovascular risk and the mortality conversation belong in the same clinic visit as the cast.

Check eGFR at the point of Charcot diagnosis, not at the next routine review
Book multidisciplinary follow-up rather than discharging back to routine foot screening
Record foot-risk status formally — it carried the strongest amputation signal here
Treat a Charcot diagnosis as a prompt to review cardiovascular risk management in full

Why it matters

It reframes Charcot from a local foot problem into a signal that the patient is at high risk of dying.

The statistics, in plain English

With only 140 patients and 44 amputations, the confidence intervals are wide: 2.05 to 35.45 for the foot-risk hazard ratio means the true effect could be modest or very large, though the interval excluding 1 makes a real association likely. This is observational — the associations identify who does badly, not what would change it. Absolute event rates are the more reliable figure here.