High vs. low glycemic index (GI) foods elicit late postprandial activation of mesolimbic brain areas, which may contribute to the associated increases in hunger, food intake and cravings. It is unknown whether changes in blood glucose or higher insulin levels mediate this activation and its potential adverse downstream effects. Fifteen men with T1D, mean (SD) BMI 21 (1.4) kg/m2, aged 21 (1.6) yr underwent three conditions in random order: C1 - high GI (GI=84) meal with IV insulin to maintain euglycemia; C2 - macronutrient matched low GI (37) meal with IV insulin to maintain euglycemia; C3 - high GI meal with IV insulin to match C2. Insulin was algorithmically adjusted for 4.5 hr with blood glucose checks every 5 min. Brain imaging (arterial spin labeling) was obtained at baseline, in the early (1 hr) and late (4 hr) postprandial period. This design allowed us to compare high vs. low GI meals with similar glucose response (incremental area under the curve 192 [22] vs. 41 [12], p=0.6) but different insulin doses (25 [2] vs. 17 [1] units, p=0.003; C1 vs. C2) or similar insulin doses (17 [1] units each, p=0.5) with different glucose response (1257 [37] vs. 41 [12] mg/dl*h, p<0.001, C3 vs. C2). As expected, a high GI meal elicits increased hunger and late postprandial activation of brain areas linked to food cravings. This activation was mediated by postprandial hyperglycemia rather than insulin level. Thus, a high GI meal may promote overeating with insulin dose deviations in either direction: 1) via mesolimbic signaling if hyperglycemia results from inadequate insulin coverage, or 2) via compensatory carbohydrate intake if excessive coverage causes hypoglycemia.


B.S. Lennerz: None. F. Munsch: None. D.C. Alsop: Other Relationship; Self; General Electric, Hitachi America, Ltd., Philips Healthcare, Siemens Corporation. G.M. Steil: Consultant; Self; Eli Lilly and Company. D.C. Simonson: Stock/Shareholder; Self; GI Windows. Stock/Shareholder; Spouse/Partner; Phase V Technologies, Inc. D.S. Ludwig: None.

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.