Objective: To model the probability of improvement on gait and cognition among patients who received intranasal insulin (INI)

Methods: 223 patients (1diabetes, 117 controls) randomized into 4 arms: 55 diabetes placebo, 51 diabetes INI, 59 control placebo, 58 control INI, 1women, 66±9 years of age. INI treatment lasted 24 weeks with 40 IU of INI (Novolin® R, off-label use) or placebo (sterile saline) . normal walk speed (NW) , dual task walk speed (DT) , executive function (EF) , verbal memory (VB) . Improvement due was defined to be better performance at end of treatment compared to baseline. Eleven of 20 baseline predictors were selected in a multivariable generalized linear model with logit link and binomial error to predict the probability of improvement among the 1patients who received INI.

Results: 58% of the INI patients improved on NW, 62% improved on DT, 69% EF, and 47% VB. The predictors for the final model included 6 baseline demographic variables (age, number of years of diabetes, number of years in school, Wechsler Adult Reading test, global depression score, and waist circumference; and 5 metabolism-related variables of hemoglobinA1c, fasting plasma glucose, total cholesterol, microalbumin in urine, and microalbumin to creatinine ratio.

Conclusion: Modeling technique also allows us to compute the estimated probability of improvement for any individual patients, thus, enabling us to understand better the particular demographic and metabolic profiles that best predict who among those receiving INI could benefit the most from the treatment. Funded by NIH-NIDDK (1R01DK103902) , FDA (IND#107690) , Novo Nordisk, Inc. (ISS-001063) , Medtronic Inc. (NERP15-0310) and registered on www.clinicaltrials.gov (NCT02415556.)


L.H.Ngo: None. V.Novak: None. P.Novak: Stock/Shareholder; Moderna, Inc., Novavax, Pfizer Inc., Prothena. C.Mantzoros: Advisory Panel; Novo Nordisk. W.Dai: None. V.Lioutas: None.


Funded by NIH-NIDDK (1R01DK103902) , FDA (IND#107690) , Novo Nordisk, Inc. (ISS-001063) , Medtronic Inc. (NERP15-0310)

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.