Obesity, defined as a body mass index (BMI) > 30 kg/m², is a growing, multifactorial condition with major implications for healthcare systems worldwide (1). Over the past four decades, the prevalence of obesity has tripled (1). Obesity is associated with numerous life-limiting comorbidities and plays a central role in the development of metabolic syndrome (2, 3).

When lifestyle modifications fail to achieve sufficient weight reduction, metabolic surgery has proven to be an effective therapeutic approach (4). Most patients maintain substantial weight loss and experience lasting improvement in obesity-related comorbidities (4). Two of the most commonly performed surgical procedures are Roux-en-Y gastric bypass (RYGB) and sleeve gastrectomy (SG) (5, 6). In addition to inducing significant weight loss, these procedures positively impact conditions such as hypertension, dyslipidemia, nonalcoholic fatty liver disease (NAFLD) or nonalcoholic steatohepatitis (NASH), sleep apnea and type 2 diabetes mellitus (T2D) (2, 3, 5).

In recent years, other combined restrictive-malabsorptive techniques—such as one-anastomosis gastric bypass (OAGB), single-anastomosis sleeve–ileal bypass (SASI), and single-anastomosis duodenal–ileal switch (SADI-S)—have gained popularity, yielding similarly favorable weight loss and glycemic control (5).

T2D is characterized by pancreatic β-cell dysfunction and reduced insulin sensitivity, leading to increased insulin demand (7). Its incidence has risen dramatically over the past decades (8–10). Approximately 80%–90% of patients with T2D are overweight or obese (8–10), and the risk of developing T2D increases with the degree of obesity (10). Weight loss improves glycemic control, and both RYGB and SG have been repeatedly compared in terms of weight reduction and remission of associated diseases (11–13). Overall, both procedures achieve similar outcomes (11–13). However, RYGB appears superior in long-term weight maintenance and discontinuation of oral antidiabetic medications (14–16), especially in patients with more severe or long-standing diabetes (17, 18). Notably, improvements in glycemic control, including normalization of glycated hemoglobin, often occur early after metabolic surgery, before significant weight loss is achieved, suggesting weight-independent metabolic effects (19, 20). The mechanisms underlying this rapid metabolic improvement remain incompletely understood.

In our previous publication, the metabolic effects of different surgical interventions were analyzed from a functional perspective, demonstrating early improvements in glucose tolerance and insulin secretion following bypass surgery (21). Building on these findings, the present study, applies a novel artificial intelligence (AI)-assisted histomorphological approach to examine pancreatic tissue in the same cohort of Zucker Diabetic Fatty (ZDF) rats, allowing a direct link between previously reported metabolic outcomes and tissue-level adaptations. Specifically, we compare duodenojejunostomy (DJOS), sleeve gastrectomy (SG), and the combined procedure (DJOS + SG), focusing on how these interventions differentially affect pancreatic structure, β-cell proliferation, differentiation, and overall histomorphology. This analysis aims to elucidate the mechanisms by which different surgical interventions influence pancreatic glucoregulation and histomorphology, thereby contributing to the distinct improvements in glycemic control observed after metabolic surgery.

By linking the previously reported functional outcomes with long-term tissue-level changes, this study provides mechanistic insight into how bariatric surgery preserves β-cell mass, enhances proliferation and differentiation, and modulates pancreatic exocrine tissue composition.

Bariatric surgery provides substantial benefits in terms of weight loss and remission of obesity-related comorbidities, particularly type 2 diabetes mellitus (T2D) (2, 22–25). In this study, we used Zucker Diabetic Fatty (ZDF) rats, a well-established monogenic model for T2D, which develops severe obesity and diabetes (26). It is important to note that this model primarily reflects a single-gene defect, whereas the human T2D condition is multifactorial and influenced by genetic, environmental, and lifestyle factors. A hallmark of ZDF rats is β-cell loss during adolescence, leading to the rapid onset of severe diabetes (26, 27).

In clinical practice, treatment strategies are increasingly shifting away from a “one-procedure-fits-all” paradigm towards more personalized approaches. Determining the optimal surgical procedure for individual patients remains a key challenge. Various recommendations have been proposed in the literature, including the use of bariatric surgery calculators to aid decision-making (28). In most cases, the initial choice is between sleeve gastrectomy (SG) and Roux-en-Y gastric bypass (RYGB) (15).

Several studies and meta-analyses have demonstrated that bariatric surgery provides superior long-term glycemic control in patients with T2D compared to drug therapy alone (2, 24, 25). However, despite initial diabetes remission, some patients experience relapse over time (25, 29). In the short term, SG has been shown to be as effective as RYGB in achieving glycemic control (11, 20, 30–33). The long-term superiority of RYGB over SG remains inconclusive, with studies reporting mixed results (11, 15, 34–39). Nevertheless, there is evidence suggesting that RYGB may confer to a potential advantage regarding sustained antidiabetic effects.

To investigate mechanisms underlying these outcomes, we employed a rat model exhibiting an extreme diabetic phenotype. Ob/ob rats, which carry a leptin receptor mutation, develop severe T2D during adolescence, eventually leading to complete endocrine pancreatic insufficiency (40). The SG group underwent substantial stomach resection to produce a restrictive effect, analogous to the human procedure. DJOS involved duodenal und proximal jejunal exclusion, analogous to human SADI-S (single anastomosis duodeno-ileal bypass with sleeve gastrectomy) or OAGB (one-anastomosis gastric bypass), without stomach resection. The combination procedure (DJOS + SG) added a restrictive component.

Importantly, the animals analyzed here represent the same cohort used in our previous publication comparing SG, DJOS, and DJOS + SG (21). Early postoperatively, all groups exhibited similar glucose tolerance; however, at 3 and 6 months, the bypass groups demonstrated superior glycemic outcomes (21). Fasting insulin levels measured six months postoperatively—a substantial timeframe for ZDF rats—similarly indicated that the bypass groups (DJOS and DJOS + SG) achieved significantly better glycemic outcomes than SG alone (Figures 2A–C) (21). These results are consistent with clinical data: McTique et al. reported that five years after RYGB, patients exhibited slightly higher T2D remission rates, fewer relapses and improved long-term glycemic control compared to SG (15).

Building on these previously published data, the current analyses extend the findings to long-term pancreatic morphology, β-cell proliferation, differentiation, and tissue composition, providing mechanistic insights that were not included in the prior study.

Given the natural β-cell loss in our model, we investigated pancreatic changes following bypass procedures. Microscopic examination revealed β-cell clustering in the bypass groups (DJOS and DJOS + SG), whereas SG animals predominantly exhibited solitary β-cells. These observations align with rodent studies showing RYGB-induced β-cell mass (41–43). Human studies similarly report improved β-cell function after bariatric surgery, as indicated by fasting glucose and insulin values, disposition index (DI), proinsulin-to-C-peptide ratio, and insulin secretion relative to glucose (44–47). Short-term outcomes, however, do not differ significantly between RYGB and SG, suggesting that mechanisms conferring RYGB's long-term advantages manifest over time (44, 46, 47). Our previously published data confirm an early glycemic benefit in SG animals that diminishes in the long term (21).

Proliferating Cell Nuclear Antigen (PCNA) and PDX-1 were used to assess β-cell proliferation and differentiation. PCNA reflects regenerative capacity, while PDX-1 is essential for β-cell function, including insulin gene expression and secretion; downregulation can lead to β-cell dedifferentiation (48, 49). Talchai et al. propose that the loss of β-cell function in T2D is due to dedifferentiation, accompanied by a decrease in maturation markers (50, 51). Notably, bariatric procedures can reverse dedifferentiation and enhance β-cell identity and functionality (52–54). In our study, bypass groups exhibited higher β-cell mass, increased proliferation markers, and elevated PDX-1 levels compared to SG alone (Figure 4D). Li et al. similarly describe a marked rise in PDX-1 levels at mRNA and protein levels in diabetic rats following RYGB surgery peaking two weeks postoperatively (55). Importantly, the SG procedure in the DJOS + SG group did not provide additional metabolic benefits, suggesting that the bypass itself drives β-cell proliferation and long-term antidiabetic effects.

Quantitative AI-based tissue analysis revealed distinct fatty degeneration and loss of acinar cells in the SG group compared to the bypass groups (Figure 5C). Pancreatic steatosis is commonly associated with obesity and metabolic disorders (56–58), and bariatric surgery has been shown to reduce pancreatic fat deposits in both rodents and humans (58, 59). Notably, Salman et al. report a positive correlation between weight loss and the reduced pancreatic fat volume following SG (58, 59). Our findings indicate that different bariatric procedures may differentially affect pancreatic exocrine tissue, with bypass surgery preserving acinar tissue and reducing fatty infiltration.

Interestingly, across all measures examined—including glucose tolerance, fasting insulin levels, markers of β-cell proliferation and function, as well as histological assessment of pancreatic steatosis and acinar tissue—the addition of a restrictive component in DJOS + SG did not confer any further benefit over DJOS alone. This suggests that the long-term metabolic improvements appear to be predominantly driven by duodenal exclusion, potentially reflecting a ceiling effect in hormonal adaptation or the dominant influence of malabsorptive, hormone-mediated mechanisms.

A limitation of this study is the restricted translatability of the animal model, given anatomical differences such as an intact pylorus and relatively large gastric volume. The model was designed to compare metabolic consequences of duodenal bypass vs. sleeve gastrectomy, rather than to replicate a specific human procedure. Additionally, the ZDF rat represents a monogenic model of T2D, whereas the disease in bariatric patients is typically multifactorial. Nonetheless, the metabolic effects observed provide valuable mechanistic insights.

Overall, these results indicate that bariatric surgery—particularly bypass procedures—beneficially impacts long-term pancreatic morphology and function. In ZDF rats, where β-cells typically undergo programmed cell death during adolescence, bypass surgery preserved β-cell proliferation and clustering even after six months, suggesting a protective effect against β-cell apoptosis.

By using the same cohort of animals previously reported by Laessle et al. (21), we demonstrated that early improvements in glycemic control are maintained long-term and are accompanied by sustained enhancements in β-cell mass, proliferation, differentiation, and pancreatic tissue composition.

Although the model cannot fully replicate human T2D, these findings support clinical observations that bypass surgery offers superior metabolic benefits at the β-cell level. These findings align with clinical evidence from McTigue et al., demonstrating that bariatric bypass surgery provides superior outcomes for T2D compared to sleeve gastrectomy (15). However, it's important to note that specific patient groups—such as those with early-stage diabetes—can benefit equally from sleeve procedures, as shown in Aminian et al., highlighting the need for individualized treatment strategies (28).

Our data demonstrate that alterations of gastrointestinal anatomy via malabsorptive or combined bariatric surgery significantly improve glycemic control in a T2D rat model. These effects are associated with preservation of β-cells, increased β-cell number and clustering, enhanced proliferation and differentiation, and histological improvements in pancreatic tissue, including increased acinar content and reduced fatty infiltration. Importantly, the animals analyzed in this study are the same cohort previously published by Laessle et al. (21), in which early metabolic improvements were reported. The current analyses extend those findings by showing that these long-term metabolic benefits are accompanied by sustained enhancements in β-cell morphology, proliferation, differentiation, and exocrine pancreatic tissue composition, providing mechanistic insight into how bariatric surgery exerts durable antidiabetic effects.