Dietary Glycemic Index, Brain Function and Food Intake in Patients With Type 1 Diabetes Mellitus

Processed carbohydrates cause rapid changes in blood sugar and have been associated with overeating and obesity. We have shown that test meals high in processed carbohydrate affect brain areas involved in addiction, craving and overeating. It is unknown whether the changes in blood sugar or the associated higher insulin levels mediate this brain activation and its likely adverse effects. Answering this question is important for patients with type 1 diabetes who have elevated risks of obesity and disordered eating: If blood sugar is the causal mechanism, optimal insulin coverage should be protective. If insulin is the causal mechanism, however, a diet high in processed carbohydrate could predispose to overeating and weight gain, as this diet requires higher insulin doses. To disentangle these factors, we will study brain activation and relevant blood markers in 15 men with diabetes. In 4 sessions, we will examine meals with differential carbohydrate properties while giving insulin infusions.

A total of 15 male participants (age 18-45) with T1DM will be recruited. Participants will be enrolled in the study for a total of 1-3 months, and participate in a pre-test visit and three test visits, each after a 10-12-hr overnight fast. Participants will be instructed to consume their regular, weight maintaining diet between visits. At the pre-test visit, the study director or PI will meet participants, confirm eligibility and obtain informed consent. Participants will receive a low glycemic index (GI) meal with optimal iv insulin coverage using a negative feedback algorithm to maintain euglycemia (euglycemic clamp). Insulin requirement will be quantified. At some time during the visit, participants will present to the BIDMC research imaging facility for a practice MRI session, during which they will undergo a brief imaging sequence to get accustomed to the scanning process and eliminate anxiety as a confounder of imaging data. At each of 3 test visits, one of the following experimental conditions will be applied in a randomized, blinded cross-over design: (a) high GI meal with euglycemic clamp, (b) low GI meal with euglycemic clamp, (c) high GI meal with primed-variable insulin infusion at the rate established during the pre-test visit. After steady state is established, baseline laboratory evaluation and MRI imaging will be obtained, followed by the test meal. Imaging will be repeated at 1 and 4 hours postprandial. Blood samples for pertinent metabolic and hormonal parameters will be obtained every 30 minutes. Each test-visit concludes with a standard weighed meal to quantify ad-libitum intake.

A nutritional shake with high GI will be consumed. Regular insulin will be administered intravenously according to a negative feedback algorithm to maintain euglycemia.This condition results in euglycemia with high insulin levels.

A nutritional shake with high GI will be consumed. Regular insulin will be administered intravenously at a rate previously established to maintain euglycemia after a low glycemic index meal. This condition results in moderate hyperglycemia with low insulin levels.

A nutritional shake with low GI will be consumed. Regular insulin will be administered intravenously according to a negative feedback algorithm to maintain euglycemia. This condition results in euglycemia with low insulin levels.

High and low GI liquid test meals are matched for macronutrient composition (60% carbohydrate, 15% protein, 25% fat), micronutrient profiles, physical properties, palatability and sweetness. Meals will provide 25% of individual daily energy requirements as estimated by the Harris Benedict equation. A high glycemic index of ~90 is achieved by using corn syrup as a carbohydrate source.

High and low GI liquid test meals are matched for macronutrient composition (60% carbohydrate, 15% protein, 25% fat), micronutrient profiles, physical properties, palatability and sweetness. Meals will provide 25% of individual daily energy requirements as estimated by the Harris Benedict equation. A low glycemic index of ~40 is achieved by using uncooked corn starch as a carbohydrate source.

Insulin will be given intravenously for 5 hours. During the entire clamp protocol, glucose levels will be measured every 5 minutes. A basal insulin infusion will be started at 80% of the patients insulin pump basal rate, and will be adjusted between 0.1 and 2.5 mU/kg•min, depending upon the patient's plasma glucose level in relation to the target range target of 90-100 mg/dl.

A primed-variable infusion of insulin will be administered at the rate established to achieve euglycemia after a low glycemic index meal. This is expected to result in moderate hyperglycemia as the high GI meal is associated with higher insulin requirements. For patient safety, glucose levels will be measured every 30 minutes. If glucose levels are > 400 mg/dl or < 60 mg/dl, insulin infusion will be adjusted to maintain glucose levels target of 60-400 mg/dl.

Cerebral blood flow in the right and left nucleus accumbent was measured by arterial spin labeling (MRI). Blood flow was normalized for whole brain perfusion and corrected for baseline perfusion in the respective brain area and meal order, as per our a priori statistical analysis plan.

Cerebral blood flow in the right and left nucleus accumbent was measured by arterial spin labeling (MRI). Blood flow was normalized for whole brain perfusion and corrected for baseline perfusion in the respective brain area and meal order, as per our a priori statistical analysis plan.

Cerebral blood flow was measured by arterial spin labeling (MRI). Grouped MRI data was visually inspected for postprandial differences between conditions. Blood flow from a cluster contracting the conditions in the right dorsal caudate, just lateral to the nucleus accumbent, was extracted, normalized for whole brain perfusion and corrected for baseline perfusion in the respective brain area and meal order, as per our a priori statistical analysis plan.

Cerebral blood flow was measured by arterial spin labeling (MRI). Grouped MRI data was visually inspected for postprandial differences between conditions. Blood flow from a cluster contracting the conditions in the right ventrolateral striatum, just lateral to the nucleus accumbent, was extracted, normalized for whole brain perfusion and corrected for baseline perfusion in the respective brain area and meal order, as per our a priori statistical analysis plan.

Functional Connectivity of Nucleus Accumbens, Hypothalamus and Other Brain Areas Involved in Intake Regulation

Cerebral blood oxygen concentration level was measured by resting state functional MRI (rs-fMRI). Seed based analysis was performed with the seed on the right Nucleus Accumbens. Functional connectivity between Nucleus Accumbens and Hypothalamus was assessed through extraction of temporal correlation measures.

Functional Connectivity of Nucleus Accumbens, Hypothalamus and Other Brain Areas Involved in Intake Regulation

Cerebral blood oxygen concentration level was measured by resting state functional MRI (rs-fMRI). Seed based analysis was performed with the seed on the right Nucleus Accumbens. Functional connectivity between Nucleus Accumbens and Hypothalamus was assessed through extraction of temporal correlation measures. Functional connectivity between Nucleus Accumbens and other brain areas was visually assessed.

Inclusion Criteria: Type 1 diabetes for a minimum of 3 years BMI 20-35 kg/m2 Use of insulin pump Willing and able to: Maintain weight and document for duration of the study Exclusion Criteria: Insulin resistance (current insulin requirement > 1.5 U/kg/d) Insulin requirement < 0.5 unit/kg/day (cut-off for preserved beta-cell function) HbA1C ≥ 8.0% DKA within 2 months Frequent hypoglycemia (BG <50 mg/dl), > 3 times per week Fluctuations in body weight >10% over preceding year Smoking or illicit substance abuse High levels of physical activity (≥60 minutes per day, ≥ 4 days per week) Current weight loss diet Medical problems, medications or dietary supplements that may affect metabolism, insulin action, body weight, appetite, energy expenditure, or gastrointestinal absorption (e.g. celiac disease) Allergies to compounds or intolerance of the liquid meals MRI exclusion criteria Other conditions according to self-report that would prohibit participation based and researcher assessment