Sirs, We read with interest a review article by De Ridder et al.1 They concluded that bariatric surgery is a promising treatment in obese patients with non-alcoholic fatty liver disease. Weight reduction appears to decrease the grade of steatosis. A recent randomized study compared the use of adjustable gastric banding to conventional treatment in obese patients with type 2 diabetes.2 Surgery was more likely to achieve remission of diabetes through greater weight loss. In the study, nearly all patients who lost greater than 10% in weight remitted. A surgical approach represents a new horizon in the treatment of obese diabetic patients and disease associated complications. Placement of an intragastric balloon may be an attractive alternative in such patients. In one study, its placement for 6 months resulted in an excess weight loss of 33.9 ± 18.7%.3 In a proof-of-concept study, we postulated that, by reducing weight and diet intake via placement of intragastric balloon, diabetic control could be improved in the obese patients with poorly controlled type 2 diabetes. Between January and June 2007, patients who fulfilled the following criteria were invited to participate in the study: body mass index (BMI) ≥ 27, poorly controlled diabetes [glycosylated haemoglobin (HbA1c) ≥ 8] for the past 3 months and age between 18–70 years. They underwent an upper endoscopy and were excluded from the study if they were found to have large hiatal hernia, severe oesophagitis, peptic ulceration and/or previous gastric surgery. Then a Bioenteric Intragastric Balloon (BIB, Inamed Health, Santa Barbara, CA, USA) was inserted according to the manufacturer’s instruction, and the balloon was inflated with 500 mL saline with methylene blue in each subject. Control subjects were those who fulfilled same inclusion criteria, but without the insertion of an intragastric balloon. They were recruited at the same period at a 2:1 ratio to the case subjects. Both groups were followed up by an endocrinologist blind to the procedure undertaken. Both groups were given same dietary and exercise advice. Both body weight and HbA1c were measured and compared at the end of 6 months in the two groups. Six patients with the median age 42 years (range 34–57 years) received placement of an intragastric balloon and 12 patients with median age 53 years (range 33–75 years) were recruited as controls. There were no significant differences in age (P = 0.075), gender (50% vs. 75% male, P = 0.34), pre-BMI (35.7 vs. 28, P = 0.16) and pre-HbA1c (8.6% vs. 9.0%, P = 0.13) in the two groups. However, after the insertion of intragastric balloon in the treatment group, the HbA1c was much lower than the control group (7.3% vs. 9.0%, P = 0.002). When assessment was made within the group with intragastric balloon inserted, both the BMI and HbA1c were improved (BMI: 35.7 vs. 33.6, P = 0.006; HbA1c: 8.6% vs. 7.3%, P = 0.007; Figure 1). But no such improvement was seen in the control group during the follow-up period (BMI: 28 vs. 28.1, P = 0.91; HbA1c: 9.0% vs. 9%, P = 0.76; Figure 1). Body mass index (BMI) and glycosylated haemoglobin (HbA1c) at baseline and at 6-month in BIB and control subjects. Pharmacological treatment of obesity has evolved as a therapy. The resultant weight loss is often less than that after bariatric surgery, which should remain the most effective way of treating type 2 diabetes in severely obese patients. It is recommended for consideration in patients with BMI ≥ 35 kg/m2.4 Pories et al.5 demonstrated that 83% of patients with diagnosed type 2 diabetes exhibited normal blood glucose and normal glycosylated haemoglobin levels 7.6 years after bariatric surgery. The Swedish Obese Subjects study reported that, after 10 years, the average weight loss from baseline was 25% after gastric bypass, 16% after vertical banded gastroplasty and 14% after gastric banding.6 The group that had undergone surgical intervention had significantly lower incidence rates of diabetes.6 However, bariatric surgery is invasive and carries a mortality rate of 0.1–4.6% and can be associated with complications such as anastomotic leakage.7, 8 On the other hand, intragastric balloon is relatively less invasive and with few complications. It is also less expensive. With this safety profile, it can also be recommended to patients with a lower BMI associated with comorbidities. Indeed, it has now been recommended for weight reduction in patients with BMI ≥ 27 kg/m2 with one obesity-related comorbidity.9 From our current pilot study, we observed a significant improvement in diabetic control in the intragastric balloon placement group, despite small sample size. We thus recommend the use of intragastric balloon as an adjunct for managing overweight patients with poorly controlled type 2 diabetes. Further large-scale randomized studies are warranted. Declaration of personal interests: None declared.