Inhibition of IL-1β Improves Fatigue in Type 2 Diabetes

Several diseases including microbial infection, rheumatoid arthritis,multiple sclerosis, and cancer have been linked to fatigue. They all have in common an upregulation of cytokines, including interleukin (IL)-1b and tumor necrosis factor-a (TNF-a), which may interfere with clock gene functions (1). Increasing evidence associates type 2 diabetes with inflammatory processes characterized by elevated production of proinflammatory cytokines and infiltration of immune cells. Reducing IL-1 activity in prediabetes and diabetes improves insulin secretion, glycemic control, andmarkers of systemic inflammation (2–4). Given this background, we hypothesized that fatigue levels may be increased in type 2 diabetes and may be improved by IL-1b antagonism. Within a placebo-controlled, doubleblind study of IL-1b antagonism with a monoclonal anti–IL-1b antibody, XOMA052, involving 30 patients with type 2 diabetes (4), we evaluated fatigue using the Fatigue Scale forMotor and Cognitive functions (5). Besides differentiating between cognitive and motor fatigue, this scale offers a subdivision into different grades of fatigue severity. At baseline, according to predefined cutoff values, 47% of the patients had no, 20% had mild, 17% had moderate, and 16% had severe fatigue, meaning that more than half of the patients suffered from considerable fatigue symptoms compared with a healthy population (5). A significant correlation between fatigue and duration of diabetes was evident (R 5 0.532, P 5 0.002). This correlation was stronger for cognitive fatigue (R 5 0.541, P 5 0.002) compared with motor fatigue (R 5 0.486, P 5 0.007). No correlation between fatigue and age, HbA1c, body weight, body temperature, and C-reactive protein was found. One month after treatment with XOMA052, a univariate ANOVA with the preand 1 month postmedication difference on total fatigue as the dependent variable and dosage as the fixed factor revealed that in the placebo and the lowest dose group (0.01 mg/kg), fatigue was slightly increased; in the two medium dose groups (0.03 mg/kg and 0.1 mg/kg), fatigue was slightly decreased; and in the two highest dose groups (0.3 mg/kg and 1.0 mg/kg), fatigue was decreased remarkably. The effect size for this dose-dependent effect was d 5 0.3. When assessing the motor and cognitive function separately, a nonparametric analysis of preand 1 month post-medication effects revealed a meaningful trend (P 5 0.07) on decrease in motor fatigue for patients under the dosage of 1.0 mg/kg of XOMA052. To further evaluate these findings with respect to the small group sizes, effect sizes for preand 1 month postmedication comparisons were calculated. Here it could be confirmed with an effect size of d5 1.05 that a dosage of 1.0 mg/kg of monoclonal anti–IL-1b antibody had a favorable effect on motor fatigue. To our knowledge, this is the first study assessing fatigue in diabetes by means of a validated fatigue instrument. It shows that type 2 diabetic patients are more prone to fatigue than normal healthy individuals with a prevalence of more than 50%. Fatigue seems to be correlated with duration of diabetic disease but not with the extent of glycemia or C-reactive protein levels. Moreover, fatigue seems to partly improve following IL-1b blockade.