Gliosis, a cellular response to injury in neural tissue, is linked to diet-induced obesity in rodent models. Growing evidence also implicates reactive gliosis and inflammation of the mediobasal hypothalamus (MBH) as a potential mechanism connecting obesity to insulin resistance. We tested for radiologic evidence of MBH gliosis in humans in relation to impaired glucose homeostasis and type 2 diabetes (T2D), independent of obesity.

Adults with obesity and 1) T2D not on insulin (N=17), 2) impaired glucose tolerance (IGT; N=21), or 3) normal glucose tolerance (NGT; N=29) by OGTT underwent magnetic resonance imaging to measure mean bilateral T2 relaxation time (ms) in the MBH and 3 control regions. Longer T2 relaxation time is a marker of gliosis.

Groups were matched for race, sex, BMI and % body fat (P=0.99), but age differed (T2D 54 ± 8 y, IGT 46 ± 11 y, NGT 45 ± 12 y; P=0.02). Group differences in T2 relaxation time varied by brain region (interaction: chi2(6) = 65, P<0.0001, adjusted for sex and age). Bonferroni-corrected post-tests confirmed that MBH, but not control region, T2 relaxation times were longer in T2D (180 ± 26 ms) than IGT (160 ± 26 ms, P<0.0001) or NGT (152 ± 17 ms, P<0.0001) groups. Among all 67 subjects, fasting glucose, 2-hr post-glucose, and hemoglobin A1c values were each positively associated with MBH, but not control region, T2 relaxation times (interactions by region all P< 0.0001, MBH all P< 0.0001, control regions all NS). Among the 50 subjects without T2D, interactions remained significant (all P<0.0001), and greater 2-hr post-glucose and hemoglobin A1c were associated with longer MBH T2 relaxation times (β = 0.22 ms, P<0.0001 and β = 14.8 ms, P=0.002, respectively) whereas a trend was present for fasting glucose concentrations (β = 0.38 ms, P=0.06).

These data provide novel evidence that the degree of inflammation and gliosis in glucose-regulating areas of the brain is related to impaired glucose homeostasis and T2D in adults with obesity.


J.L. Rosenbaum: None. L.E. Sewaybricker: None. S. Chandrasekaran: None. M. De Leon: None. M. Webb: None. S.J. Melhorn: None. E. Schur: None.


American Diabetes Association (1-17-ICTS-085 to E.S.)

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