Memory Aid by Intranasal Insulin in Diabetes (MemAID)

The investigators propose a randomized controlled trial determining the long-term effects of
intranasal insulin (INI) on cognition and memory in type 2 diabetes (DM) and non-DM groups.
The investigators hypothesize that: 1) INI-treated adults with DM have better memory and
functioning of specific cognitive domains and faster walking during a dual task than those
treated with placebo and the control group; 2) Glycemic and insulin resistance and genetic
markers for Alzheimer's disease (Apolipoprotein E4 [ApoE4]) may serve as predictors of
positive responses to INI therapy; 3) INI treatment neither adversely affects systemic
glycemic levels or the cardiovascular system nor causes weight gain.

Aim 1: To determine whether INI-treated type 2 DM adults have a) better memory and
functioning of specific cognitive domains and b) faster dual-task gait speed and better daily
living functioning than the placebo-treated and non-DM groups. Four groups will be tested: 60
DM subjects treated with insulin; 60 DM subjects treated with placebo; 45 control subjects
treated with INI and 45 control subjects treated with placebo. These 210 patients are
expected to complete treatment and 168 are expected to complete study by the study completion
anticipated date.

The investigators will conduct a randomized, double-blind, placebo-controlled study in 120
older adults with type 2 DM and 90 non-DM controls examining whether 40 IU INI once daily
over a 24-week period improves:

- Specific domains of visuospatial attention and memory, verbal learning (primary
outcomes);

- Gait speed during a dual task (which is an excellent predictor of overall health), daily
living functionality, and depression as compared to the DM group receiving sterile
saline and the non-DM groups. The non-DM groups will provide reference of INI effects in
a clinical phenotype of cognitive decline and insulin resistance that occurs with normal
aging.

Aim 2: To identify a phenotype and long-term trajectory predicting clinically relevant
response to INI therapy based on glycemic control, insulin resistance, endothelial and
genetic markers.

1. The investigators will determine a phenotype predicting a clinically relevant response
to INI therapy and identify time-dependent trajectories of INI effects on cognition in
the DM group vs. the placebo and the non-DM groups. Clinical predictors will be based on
associations between cognitive function and/or gait and demographic, glycemic control,
insulin resistance, endothelial and genetic (ApoE4) measures.

2. The investigators will evaluate the dose-escalating trajectory of cognition, gait speed,
and functionality during the 24 weeks of therapy and 24 weeks post-treatment and their
dependence on the above-mentioned factors, and determine the time point when maximum
effect was reached. INI therapy response is defined as a clinically relevant improvement
on cognitive tests or in gait speed (as a continuous variable) or as responders vs.
non-responders as compared to placebo within DM and non-DM groups (as a categorical
variable).

3. MRI substudy: The investigators will explore the long-term INI effects on regional
perfusion, vasodilatation, and resting functional connectivity in 40 DM subjects pre-
and post- INI/placebo administration at the beginning and at the end of intervention and
their relationships to cognitive outcomes. Regional perfusion and vasodilatation will be
measured by pseudo-continuous arterial spin labeling (PCASL) MRI at 3 Tesla, and
resting-state functional connectivity will be quantified from low-frequency (0.01-0.08
Hz) fluctuations (LFF) of the whole-brain blood-oxygen-level dependent (BOLD) functional
Magnetic Resonance Imaging (fMRI).

Aim 3: To determine the long-term safety of INI vs. placebo with regard to glycemic control
(fasting glucose, hemoglobin A1c [HbA1c], hypoglycemic episodes), vital signs, and body mass.

1. The investigators will obtain measurements of fasting glucose, insulin, vital signs, and
body mass at baseline, 2-months, 4-months, and 6-months follow-up and keep weekly logs
monitoring glucose and adverse events.

2. Safety substudy: In the first 20 DM patients treated with subcutaneous insulin, the
investigators will conduct continuous glucose monitoring (CGM) OR 5 finger sticks/day
(effective after 9/25/2017) for 1 week during baseline and during the first week of INI
or placebo treatment to evaluate the INI effects on glycemic control, hypoglycemic
episodes, and body weight.

This study may pave the way to potential treatment and/or cure of DM- and age-related
cognitive decline.

Inclusion Criteria:

- Men and women aged 50-85 years old

- Able to walk for 6 minutes

- Diabetes type 2 (DM) group: diagnosis and treatment for type 2 DM with non-insulin
oral or injectable agents

- Non-DM group with similar age range as the DM group, non-diabetic fasting plasma
glucose (<126 mg/dL) and hemoglobin A1c (HbA1c) (<6.5%)

- Participants capable of providing informed consent

Exclusion Criteria:

- Type 2 DM treated wiyh insulin (as of 9/25/2017)

- Type 1 DM

- Intolerance to insulin

- History of severe hypoglycemia

- Participants who have >1 asymptomatic and/or symptomatic episode of hypoglycemia
(glucose < 70 mg/dL) during finger stick or plasma glucose

- Acute medical condition that required either hospitalization or surgery within the
past 6 months (e.g., severe hypoglycemia, malignancies, myocardial infarction,stroke)

- Liver or renal failure or transplant

- Dementia (Mini Mental State Examination [MMSE] scores ≤20)

- Current recreational drug or alcohol abuse

- Serious systemic disease that would interfere with conduction of clinical trial (mild
forms of neurological conditions e.g. Parkinson's Disease, autonomic neuropathy etc.
would be allowed)

- Magnetic Resonance Imaging (MRI) substudy in 40 DM patients only: claustrophobia and
implants incompatible with 3-Tesla MRI

- Safety substudy in 20 IDDM patients only: Insulin-treated type 2 diabetics with a
C-peptide of <0.8 ng/mLd and fasting blood glucose >150 mg/dL will be excluded even
without history of hypoglycemia during finger stick measurements.