A randomized, placebo-controlled, double-blinded trial of duloxetine in the treatment of general fatigue in patients with chronic fatigue syndrome.

OBJECTIVE: To assess the efficacy and safety of duloxetine in patients with chronic fatigue syndrome. METHODS: A 12-week, randomized, double-blind study was designed to compare duloxetine 60-120 mg/d (n = 30) with placebo (n = 30) for efficacy and safety in the treatment of patients with chronic fatigue syndrome. The primary outcome measure was the Multidimensional Fatigue Inventory general fatigue subscale (range: 4-20, with higher scores indicating greater fatigue). Secondary measures were the remaining Multidimensional Fatigue Inventory subscales, Brief Pain Inventory, Medical Outcomes Study Short Form-36, Hospital Anxiety and Depression Scale, Centers for Disease Control and Prevention Symptom Inventory, Patient Global Impression of Improvement, and Clinical Global Impression of Severity. The primary analysis of efficacy for continuous variables was a longitudinal analysis of the intent-to-treat sample, with treatment-by-time interaction as the measure of effect. RESULTS: The improvement in the Multidimensional Fatigue Inventory general fatigue scores for the duloxetine group was not significantly greater than for the placebo group (P = 0.23; estimated difference between groups at week 12 = -1.0 [95% CI: -2.8, 0.7]). The duloxetine group was significantly superior to the placebo group on the Multidimensional Fatigue Inventory mental fatigue score, Brief Pain Inventory average pain severity and interference scores, Short Form-36 bodily pain domain, and Clinical Global Impression of Severity score. Duloxetine was generally well tolerated. CONCLUSION: The primary efficacy measure of general fatigue did not significantly improve with duloxetine when compared with placebo. Significant improvement in secondary measures of mental fatigue, pain, and global measure of severity suggests that duloxetine may be efficacious for some chronic fatigue syndrome symptom domains, but larger controlled trials are needed to confirm these results.