The aim of this study was to evaluate decrease in waist circumference in obese patients receiving different anti-obesity treatments. The study was designed as a short-term (12 weeks), open-label, and randomized trial. Eighty six patients (70 females, 81.4%; mean age 41.09+/-8.73 years, mean BMI 36.1+/-4.3 kg/m(2)) were randomized to four different therapy groups. The primary outcome parameters were waist circumference and body mass index (BMI). The therapy groups were a) diet+ sibutramine 1x10 mg/d (n=22), b) diet+orlistat 3x 120 mg/d (n=25), c) combination of diet+sibutramine+orlistat (n=20) and d) diet (n=19). Combination therapy was more effective than diet and orlistat mono-therapy (p<0.0001 for all), but not significantly superior to sibutramine mono-therapy (p=0.072) in decreasing BMI. Sibutramine mono-therapy was significantly more effective in inducing BMI decrease compared with orlistat mono-therapy (p=0.039). The association between change in BMI and change in waist circumference was strongest in the orlistat mono-therapy group (P interaction=0.003). This means that patients taking orlistat experienced more decrease in waist circumference (3.4 cm, R-2=0.29) per unit decrease in BMI compared to patients under combination therapy (2.6 cm, R-2=0.25, P interaction = 0.015) and patients taking sibutramine (1.8 cm, R-2=0.19, P interaction=0.026). In the diet therapy group decline in waist circumference was independent of BMI (1.9 cm, R-2=0.02, P interaction=0.076). Although combination therapy and sibutramine mono-therapy were more effective in decreasing BMI, reduction in waist circumference and BMI was most significantly associated with the orlistat mono-therapy group. This may hint at the possibility of orlistat inducing weight loss mainly in the abdominal area targeted to reduce cardiovascular risk. (C) 2004 Tohoku University Medical Press.