Calorie restriction reduces the influence of glucoregulatory dysfunction on regional brain volume in aged rhesus monkeys

Abstract

Insulin signaling dysregulation is related to neural atrophy in hippocampus and other areas affected by neurovascular and neurodegenerative disorders. It is not known if long-term calorie restriction (CR) can ameliorate this relationship through improved insulin signaling or if such an effect might influence task learning and performance. To model this hypothesis, magnetic resonance imaging was conducted on 27 CR and 17 control rhesus monkeys aged 19-31 years from a longitudinal study. Voxel-based regression analyses were used to associate insulin sensitivity with brain volume and microstructure cross-sectionally. Monkey motor assessment panel (mMAP) performance was used as a measure of task performance. CR improved glucoregulation parameters and related indices. Higher insulin sensitivity predicted more gray matter in parietal and frontal cortices across groups. An insulin sensitivity × dietary condition interaction indicated that CR animals had more gray matter in hippocampus and other areas per unit increase relative to controls, suggesting a beneficial effect. Finally, bilateral hippocampal volume adjusted by insulin sensitivity, but not volume itself, was significantly associated with mMAP learning and performance. These results suggest that CR improves glucose regulation and may positively influence specific brain regions and at least motor task performance. Additional studies are warranted to validate these relationships.

FIG. 1.

The relationship between S_I and regional GM volume across subjects and an S_I × dietary condition interaction testing such an association for each dietary group. S_I near the time of scan was not available for one control and one CR monkey. Sixteen controls and 26 CR monkeys were analyzed. Higher S_I for both control and CR monkeys (yellow-orange color) corresponded to more GM in parietal and frontal cortices (A–C). For the interaction, CR animals showed more GM per unit increase in S_I compared with controls in hippocampus, temporal pole, agranular insula, striatum, prefrontal cortices, and anterior cingulate cortex (purple color, A–F). To illustrate these results, a representative voxel from each analysis was graphed, both for the association of S_I across groups (G) and the interaction between groups (H). A sagittal view of the brain depicts the location of the coronal slices in A–F along a posterior-anterior axis. Color bars and color maps represent t values. GM volume is depicted in arbitrary units (A.U.). Brains are oriented such that the left hemisphere is on the left side.

FIG. 2.

Partial regression plots depicting error variance of the S_I × dietary condition interaction explained by the HOMA-IR, basal insulin, and basal glucose. Control (n = 16) and CR (n = 26) monkeys are, respectively, represented by red triangles and blue circles. The green arrows indicate a representative control subject across parameters. Among controls, increases in HOMA-IR and basal glucose predicted less GM, whereas higher insulin levels were associated with more GM. These relationships were not significant among CR monkeys. GM volume is depicted in arbitrary units (A.U.).