Gastrectomy, particularly sleeve gastrectomy (SG), has been shown to be an effective surgical option for morbid obesity due to its low complication rates and significant weight loss results [1-5]. SG results in alteration of the appetite through the regulation of gut hormones, resulting in decreased hunger and increased satiety [6]. SG also improves glucose homeostasis through resulting changes in gut hormone levels [7]. Specifically, laparoscopic SG results in significant improvement in glucose metabolism in patients who are morbidly obese and has been found to stop the development of diabetes at a high rate [8]. SG has been shown to improve blood glucose independently of weight loss by restoring hepatic insulin sensitivity [9]. However, the effects of gastrectomy on patients who are not obese with type 2 diabetes mellitus (DM) are less clear, with some studies suggesting that gastrectomy may improve diabetic status [10].

In patients with gastric cancer diagnosed with type 2 DM, gastrectomy has been found to have a positive impact on their glycemic control. Improvements in glycemic control, or even diabetes remission, have been reported after gastrectomy [10-15]. The extent of the gastrectomy, duration of diabetes, and method of reconstruction have been identified as important factors influencing the improvements in glycemic control [10-14]. Although the mechanisms underlying these effects are not fully understood, oncometabolic surgeries, including gastrectomy, have been suggested as a potential treatment for type 2 DM in patients with gastric cancer [16].

Studies have shown that total gastrectomy (TG) is associated with improved glucose metabolism in patients with gastric cancer, resulting in a lower rate of newly diagnosed diabetes after surgery [17]. However, the effects of gastrectomy on glucose metabolism in patients with and without diabetes have been inconsistent, with some studies reporting significant reductions in fasting blood glucose levels after gastrectomy [18]. Furthermore, SG has been associated with significant reductions in hemoglobin A_1c levels in patients without diabetes, suggesting its possible role in the prevention of type 2 DM [19].

In terms of reconstruction after partial gastrectomy in patients with gastric cancer, both Roux-en-Y (RY) and Billroth II reconstructions have been considered acceptable options [20]. RY reconstruction is often preferred for patients with gastric cancer, given that this procedure can lead to decreased reflux gastritis and esophagitis, decreased probability of cancer recurrence, and decreased incidence of surgical complications [21]. RY reconstruction has also been found to be as effective as other methods with respect to nutritional status and postoperative outcomes [22]. In comparison to Billroth II reconstruction, RY has been shown to have similar postoperative complications and better long-term outcomes [23]. Furthermore, RY reconstruction without cutting has been the preferred method in cases of gastritis, bile reflux, and gastric residuals [24].

Various studies have examined the impact of different reconstructive procedures on postoperative complications in patients with gastric cancer. It has been found that long-limb RY bypass reconstruction could lead to improved glycemic control [25], and it has been observed that preexisting DM is associated with postoperative complications [10,26]. Several studies further support the benefits of RY reconstruction, with some indicating it to be more effective than Billroth II reconstruction [27,28]. Additionally, significant improvements in DM control have been associated with RY reconstruction [25,28,29].

Given these, the aim of this study was to investigate the incidence of new-onset DM in patients with gastric cancer after surgery and how this incidence varies with different types of surgical reconstruction, namely, the RY procedure and other alternative reconstruction techniques. While studies have investigated how surgical treatment for gastric cancer affects existing DM [25,28,29], none have investigated the development of new-onset DM in patients without DM; to the best of our knowledge, this is the very first study to do so. Findings from this study could contribute valuable insights into the postoperative outcomes associated with different gastric reconstruction techniques. Such insights are vital for guiding clinical decisions and optimizing patient care, particularly in the context of mitigating the risk of developing DM after gastric surgery. Moreover, findings from this study are expected to have significant implications for both clinical practice and future research in the field of gastric surgery and DM prevention.

The characteristics of the patients included in the study at the time of the surgery are shown in Table 1. Of the 715 patients who had a gastrectomy, 489 (68.4%) underwent RY reconstruction.

The Kaplan-Meier curve of new-onset DM in the RY and OT groups is shown in Figure 1. Granular comparison of the incidence rates of postoperative diabetes associated with these distinct reconstructive procedures was made. The rate of diabetes onset was inferred from the slope of these curves, with a steeper decline indicating a higher incidence within the respective group.

A log-rank test revealed that the hazard ratio was 1.52 (95% CI 1.06-2.18), and the resultant P value from this log-rank test was .02, which denotes a statistically significant difference in the incidence of new-onset diabetes after surgery between patients with gastric cancer who underwent RY reconstruction versus OT. These findings indicate a difference in the incidence of postoperative diabetes based on the type of gastric reconstruction method used (Figure 1).

A Swimmer plot was then produced (Figure 2). In the Swimmer plot, orange lines represent RY cases, blue lines represent OT cases, a cross indicates censoring due to death, and a circle represent censoring due to new-onset DM. The last DM onset in the OT group was at approximately 2400 days, which explains the linear part of the Kaplan-Meier curve.

Propensity score matching was conducted according to the use of laparoscopy, age, sex, and BMI. After propensity score matching, the Kaplan-Meier onset curve showed a hazard ratio of 1.42 (95% CI 1.09-1.86), and the resultant P value was .009 (Figure 3). This means that the results are robust even when accounting for unknown confounding and that RY cases have more postoperative DM than OT cases.

A Kaplan-Meier curve stratified by sex was also generated (Figure 4). In this Kaplan-Meier curve, there was no significant difference in the development of postoperative DM between the RY and OT groups for both male and female patients (P=.12 and P=.24, respectively).