A new model to estimate insulin resistance via clinical parameters in adults with type 1 diabetes

Insulin resistance (IR) is a risk factor to assess the development of micro- and macro-vascular complications in type 1 diabetes (T1D). However, diabetes management in adults with T1D is limited by the difficulty of lacking simple and reliable methods to estimate insulin resistance. The aim of this study was to develop a new model to estimate IR via clinical parameters in adults with T1D.

A total of 36 adults with adulthood onset T1D (n = 20) or childhood onset T1D (n = 16) were recruited by quota sampling. After an overnight insulin infusion to stabilize the blood glucose at 5.6 to 7.8 mmol/L, they underwent a 180-minute euglycemic-hyperinsulinemic clamp. Glucose disposal rate (GDR, mg kg^-1  min^-1 ) was calculated by data collected from the last 30 minutes during the test. Demographic factors (age, sex, and diabetes duration) and metabolic parameters (blood pressure, glycated hemoglobin A_1c [HbA_1c ], waist to hip ratio [WHR], and lipids) were collected to evaluate insulin resistance. Then, age at diabetes onset and clinical parameters were used to develop a model to estimate lnGDR by stepwise linear regression.

From the stepwise process, a best model to estimate insulin resistance was generated, including HbA_1c , diastolic blood pressure, and WHR. Age at diabetes onset did not enter any of the models. We proposed the following new model to estimate IR as in GDR for adults with T1D: lnGDR = 4.964 - 0.121 × HbA_1c (%) - 0.012 × diastolic blood pressure (mmHg) - 1.409 × WHR, (adjusted R²  = 0.616, P < .01).

Insulin resistance in adults living with T1D can be estimated using routinely collected clinical parameters. This simple model provides a potential tool for estimating IR in large-scale epidemiological studies of adults with T1D regardless of age at onset.