Changes in diabetes-related complications in the United States, 1990-2010.
Level 3 - non-randomized controlled study
Population-based trend analysis using national surveillance and registry data.
PubMed 24738668 · doi:10.1056/NEJMoa1310799
What was done
Researchers analyzed trends in diabetes-related complications in the United States from 1990 through 2010 using data from the National Health Interview Survey, National Hospital Discharge Survey, U.S. Renal Data System, and National Vital Statistics System. Incidence rates for lower-extremity amputation, end-stage renal disease, acute myocardial infarction, stroke, and death from hyperglycemic crisis were calculated, with age standardization to the 2000 U.S. population.
What was found
Between 1990 and 2010, relative rates of all five complications among adults with diabetes declined: - Acute myocardial infarction declined by 67.8% (95% CI, -76.2 to -59.3), representing an absolute reduction of 95.6 cases per 10,000 persons (95% CI, 76.6 to 114.6). - Death from hyperglycemic crisis declined by 64.4% (95% CI, -68.0 to -60.9), an absolute reduction of 2.7 deaths per 10,000 persons (95% CI, -2.4 to -3.0). - Stroke declined by 52.7% and lower-extremity amputations declined by 51.4%. - End-stage renal disease had the smallest relative drop at -28.3% (95% CI, -34.6 to -21.6). Rate reductions were greater in adults with diabetes than in those without diabetes. When examining the entire population, rates for acute myocardial infarction and hyperglycemic deaths fell by 2.7 and 0.1 per 10,000 respectively, while population-level rates of amputation, stroke, and end-stage renal disease did not decline.
Why it matters
The findings demonstrate substantial progress in diabetes clinical management and preventive care over two decades. However, because overall diabetes prevalence grew during this timeframe, the total public health burden of diabetes complications remains large.
Limits
The abstract relies on aggregate secondary administrative and survey data without specifying individual sample sizes. Residual confounding, changes in diagnostic criteria or hospital coding practices over the 20-year period, and lack of differentiation between type 1 and type 2 diabetes were not described in the abstract.
Cited by
- supports Over the last 20 to 30 years in the United States, the rate of progression of diabetic neuropathy has slowed.