In the Journals Plus

Kidney disease ‘virtually universal’ at long durations of type 1 diabetes

Show Citation

September 22, 2017

Tina Costacou
Tina Costacou

Patients with type 1 diabetes duration of 50 years are more likely than not to experience some form of kidney disease, and although incidence of end-stage renal disease has declined, incidences of microalbuminuria and macroalbuminuria have not, according to an analysis published in Diabetes Care.

“Our findings suggest that current methods of type 1 diabetes management, which largely focus on glycemic control, do not prevent the development of kidney disease, although they appear to delay kidney failure,” Tina Costacou, PhD, associate professor of epidemiology at the University of Pittsburgh, told Endocrine Today. “We must, therefore, expand our focus to intensively manage other known kidney disease risk factors (eg, cholesterol, blood pressure) as well and encourage research projects aiming to identify currently unknown risk factors, since they may offer an additional opportunity to reduce the development of kidney disease.”

Costacou and Trevor J. Orchard, MBBCh, MMedSci, FAHA, FACE, also a professor of epidemiology at the University of Pittsburgh, examined 50-year cumulative kidney complication risk in a cohort with childhood-onset type 1 diabetes diagnosed between 1950 and 1980 (n = 932; mean baseline age, 27 years; mean diabetes duration, 19 years; mean age at diabetes onset, 8 years), using data from the Pittsburgh Epidemiology of Diabetes Complications Study. Researchers assessed the cumulative incidence of microalbuminuria and macroalbuminuria among patients who completed biennial surveys and exams (n = 658); to reduce the effect of survival bias, analyses pertaining to end-stage renal disease (defined as starting kidney dialysis or undergoing kidney transplantation) included those who provided only survey data (n = 130) and those who died before study baseline for whom a death certificate was available (n = 144). Researchers estimated cumulative incidence at 10-year intervals between 20 and 50 years duration and compared by calendar year of diabetes onset.

Cumulative incidence of microalbuminuria, macroalbuminuria, ESRD, and ESRD and all-cause mortality increased with longer diabetes duration. By 20 years’ duration, more than half the cohort had microalbuminuria, increasing to 82% by 40 years. Macroalbuminuria incidence rose from 27% at 20 years’ duration to nearly 57% by 40 years’ duration. At 50 years’ duration, 61% of patients had ESRD.

In comparing patients diagnosed after 1965 with patients diagnosed between 1950 and 1965, researchers observed only a decline in ESRD — a 45% decrease by 40 years of diabetes duration. There were no declines in incidences of microalbuminuria or macroalbuminuria.

Onset before age 6 years was associated with the lowest risk, according to researchers, and incidence generally did not differ by sex.

“Some degree of kidney disease in [type 1 diabetes] is virtually universal at long durations and not declining, which has major implications for health care and research strategies,” Costacou and Orchard wrote.

Additionally, Costacou said, the number of people being diagnosed with type 1 diabetes is increasing globally, and as the life expectancy of this population increases due to improvements over time in disease management, the number affected by advanced kidney disease will also rise.

“It is, therefore, crucial that we are able to focus on prevention efforts and decrease the development of even early signs of kidney disease,” Costacou said. – by Regina Schaffer

For more information:

Tina Costacou, PhD, can be reached at the University of Pittsburgh, Department of Epidemiology, Diabetes and Lipid Research Clinic, 3512 Fifth Avenue, Pittsburgh, PA 15261; email:

Disclosures: The authors report no relevant financial disclosures.

Editor’s Note: On Sept. 22, we corrected data in the fourth and sixth paragraphs to clarify that the incidence reported was at 40 years and to reflect the correct incidence. Editors regret this error.