It is reported that body weight loss is mainly caused by loss of body fat . However, the effects of gastrectomy procedures on postoperative visceral fat have not been thoroughly examined. In this study, we analyzed the effect of surgical procedure on VFA. In the early postoperative period, fat volumes are seriously affected by the metabolic abnormality caused by surgery. Previous studies indicated that body weight loss occurs mainly during the first 3 months after surgery . Therefore, in this study, we compared CT images taken at least 6 months after surgery, when the nutritional status is presumed to be stable .
A few studies examined changes in total fat or visceral fat after gastrectomy. Liedman et al.  studied changes in body composition by measuring total body potassium and water in 75 patients with gastric cancer. They observed a significant decrease in total body fat at 6 months after surgery but a subsequent marginal recovery at 12 months. In another study, Kiyama et al.  evaluated fat mass by multifrequency bioelectrical impedance analysis before and after gastrectomy. They found a larger decrease in body fat mass after TG than after subtotal gastrectomy and laparoscopy-assisted gastrectomy. Yoon et al.  reported a greater loss of visceral fat tissue after TG than after subtotal gastrectomy at 6 months after surgery. At 12 months after surgery, the losses in BMI, total adipose tissue, subcutaneous, and visceral fat tissues were all greater after TG than after subtotal gastrectomy. Miyato et al. reported that postoperative visceral fat of the RY group tended to be smaller than that of the BI group, although the difference was not statistically significant, perhaps because of the small sample size. Based on these differences, there is a possibility that the reconstruction procedure might have some effects on VFA.
The reasons for the differences in visceral fat reduction between BI and RY could not be elucidated in this retrospective study. Previous studies on bariatric surgery are of some help. It is reported that visceral fat reduction was greater after RY gastric bypass than after vertical banded gastroplasty . Although the mechanisms of fat reduction or improvement in insulin resistance are not completely understood in bariatric surgery, the importance of gut hormones has been reported. Among the various gut hormones, gastric inhibitory polypeptide (GIP) and glucagon-like peptide-1 (GLP-1) are reported to regulate fat metabolism. GIP is released from the duodenal endocrine K cells immediately after the absorption of fat or glucose . Furthermore, fat intake induces hypersecretion of GIP, which increases nutrient uptake and triglyceride accumulation in adipocytes . In the context of this study, Korner et al.  reported lower GIP levels after RY gastric bypass than after adjustable gastric banding, and concluded that the blunted GIP secretion after RY seems to contribute to the greater weight loss and improved glucose homeostasis compared with adjustable gastric banding. GLP-1 is a naturally occurring incretin hormone with a potent blood-glucose lowering action only during hyperglycemia; a GLP-1 analogue reduces visceral fat . Peterli et al.  reported higher GLP-1 levels after RY gastric bypass than after sleeve gastrectomy. The GIP level might also be lower and the GLP-1 level might be higher after RY than BI, and fat accumulation is lower after RY than BI. Fat malabsorption might be another reason; since clinical tests after RY showed significantly lower fat absorption than after BI and double-tract reconstruction, which accommodated for the passage of food through the duodenum .
As for omentectomy, our data showed that it was not a significant determinant of visceral fat loss by multivariate analysis. There are some reports of surgical removal of visceral fat thorough omentectomy [25, 26]. There was no significant difference between RY gastric bypass without omentectomy and with omentectomy in fat mass 12 months after surgery ; this is compatible with our result.
These results might be helpful in selecting the reconstruction method used after distal gastrectomy, especially for obese patients when remnant stomachs are large enough so that either BI or RY could be performed. However, our study has several limitations. First, a retrospective sample was used in this analysis, and thus the clinical background, such as extent of lymph node dissection, omentectomy, and surgical approach, is different between compared groups, although these factors did not correlate with visceral fat loss in this study. Next, the size of remnant stomach, which could affect the amount of food intake after surgery, might be smaller after RY than after BI, because the former construction technique was often selected when the remnant stomach was too small to allow for anastomosis with the duodenum. Third, the long-term results are unknown, because we examined CT data only 1 year after surgery.
Further prospective studies with stratified randomization and long-term follow-up data, such as VFA after 3 years or 5 years and cause of death, are needed to confirm the effects of reconstruction after gastrectomy on visceral fat.