Effect of bariatric surgery timing on breast cancer incidence: Real-world evidence from TriNetX Authors: Kian-Hwee Chong, Yao-Jen Chang, Joshua Wang, Kuo-Cheng Lu, Tsai-Chieh Hou, Ching-Feng Cheng, Kuo-Wang Tsai Published: 2026 ABSTRACT PurposeObesity increases the risk of breast cancer (BC), and bariatric surgery may lower overall cancer risk. However, the impact of sleeve gastrectomy (SG) timing on cancer outcomes remains unclear.Patients and methodsUsing the TriNetX global health research network, we identified 4,439,072 women diagnosed with severe obesity (BMI >35 kg/m2) between 2010 and 2019. Patients were grouped into SG and nonsurgical controls, with SG patients further stratified by surgery timing: early (≤6 months), intermediate (6–12 months), and delayed (1–3 years) post-diagnosis. Propensity score matching adjusted for demographics, comorbidities, BMI, and HbA1c. Outcomes included 5-year incidences of mortality, cancer, and metastasis.ResultsSG was associated with lower 5-year all-cause mortality across all timing groups compared to the control group. Early SG was significantly associated with a reduced 5-year incidence of BC (risk ratio [RR]: 0.799; 95% CI: 0.647 to 0.987; P = .037), gynecologic cancer (RR: 0.490; 95% CI: 0.378 to 0.635; P < .001), and metastasis (RR: 0.306; 95% CI: 0.230 to 0.408; P < .001). By contrast, delayed SG was associated with an increased risk of BC (RR: 1.404; 95% CI: 1.107 to 1.781; P = .005). Age-stratified analyses indicated that the association between SG timing and outcomes appeared more pronounced in patients aged ≥45 years.ConclusionsOur findings highlight the importance of timely bariatric surgery in optimizing both metabolic and oncologic outcomes, particularly in older women with obesity. Introduction Breast cancer (BC), the most frequently diagnosed malignancy in women worldwide, remains a major global health challenge [1]. The increasing prevalence of obesity in recent decades has further complicated this challenge, given the strong positive association between obesity and BC risk [[2], [3], [4], [5]]. Specifically, obesity influences cancer progression: cancer-associated adipocytes actively participate in tumor growth, metastasis, and survival [[6], [7], [8], [9]]. Bariatric surgery, an effective intervention for achieving substantial and sustained weight loss, has been associated with a reduced risk of BC [[10], [11], [12]]. Sleeve gastrectomy (SG) is the most commonly performed bariatric surgery worldwide, accounting for 55.4% of all bariatric interventions in 2018 [13]. SG has gained popularity because it has efficacy similar to and complication rates lower than those of other procedures, such as Roux-en-Y gastric bypass [14]. Given the biological consequences of prolonged obesity-related metabolic and hormonal changes, the timing of bariatric surgery may modulate the risk of BC. However, most studies investigating how surgery timing influences BC risk have grouped various bariatric interventions together, which obscure the individual effect of SG on BC risk and the role of surgery timing in cancer prevention. Very few studies have focused on the effect of surgery timing on BC incidence. Specifically, it remains unclear whether there is an association between the preoperative duration of obesity, the timing of SG, and the risk of BC. Thus, rigorous and focused research is required to investigate the association between SG timing and BC incidence. In the present study, we analyzed data from the TriNetX database to explore the associations between preoperative obesity duration, SG timing, and the risk of BC. Our findings may guide clinical decision-making regarding effective bariatric surgery strategies for cancer prevention. Methods Study design and data source This retrospective observational cohort study was conducted using real-world data from TriNetX, a global federated research network (Cambridge, MA, USA). This database analyses deidentified electronic health records from numerous health-care organizations (eg, academic medical centers, community hospitals, and specialty clinics) across multiple countries. TriNetX offers a secure platform for researchers to query and analyze aggregated patient data structured according to standardized billing codes, such as demographic characteristics, diagnoses (International Classification of Diseases, Tenth Revision [ICD-10] codes), procedures (Current Procedural Terminology codes), laboratory results (Logical Observation Identifiers Names and Codes), medications (Anatomical Therapeutic Chemical classification), and mortality-related information. Women with obesity were selected on the basis of predefined selection criteria. Data collection was completed on May 30, 2025. The study protocol was approved by the Institutional Review Board of Taipei Tzu Chi Hospital, Taiwan (approval number: 14-IRB015). Study cohort The study cohort comprised women (age: 18 to 90 years) with a body mass index (BMI) of >35 kg/m2 or a confirmed diagnosis of severe obesity (identified using relevant ICD-10 codes; eg, E66.0 to E66.9). Fulfilling either of these criteria is now referred to as receiving a diagnosis of severe obesity. To ensure an adequate follow-up period for assessing 5-year outcomes, we only included patients who received a severe obesity diagnosis between January 1, 2000, and May 14, 2019. The included patients were divided into 2 groups: SG group and nonsurgical (control) group. The SG group women received a diagnosis of severe obesity and subsequently underwent SG (identified using specific Current Procedural Terminology codes [eg, 43644, 43775, 43845, and 43770] or equivalent codes for laparoscopic SG). The control group comprised women who received a diagnosis of severe obesity but did not undergo bariatric surgery during the study period. Patients who underwent multiple gastric reduction procedures were excluded from this study. To investigate the effects of SG timing (relative to the first diagnosis of severe obesity) on 5-year outcomes, the SG group was further stratified by surgery timing into 3 subgroups: early intervention subgroup (SG within 6 months from diagnosis), intermediate intervention subgroup (SG within 6 to 12 months from diagnosis), and delayed intervention subgroup (SG within 1 to 3 years from diagnosis) (Fig. 1A; Supplementary Fig. 1A-1D). These timing categories were prespecified prior to outcome analysis based on real-world data indicating that bariatric surgery wait times frequently range from several months to more than one year, with reported median wait times of approximately 3–5 months and a substantial proportion of patients waiting longer than 6 or even 12 months [15,16]. The ≤6-month, 6–12-month, and 1–3-year intervals were therefore selected to represent early, intermediate, and prolonged surgical delay. The upper limit of 3 years was chosen to capture extended postponement while maintaining adequate follow-up for 5-year outcome assessment.Fig. 1Flowchart Depicting Patient Selection and Group Allocation.Eligible patients were included on the basis of predefined selection criteria. (A) Patients who had undergone SG were stratified into 3 groups by surgery timing (relative to the first recorded diagnosis of severe obesity). (B) The entire cohort was divided into either SG or control group. After propensity score matching, 5-year outcomes were analyzed. Abbreviation: SG, sleeve gastrectomy. Flowchart Depicting Patient Selection and Group Allocation. Eligible patients were included on the basis of predefined selection criteria. (A) Patients who had undergone SG were stratified into 3 groups by surgery timing (relative to the first recorded diagnosis of severe obesity). (B) The entire cohort was divided into either SG or control group. After propensity score matching, 5-year outcomes were analyzed. Abbreviation: SG, sleeve gastrectomy. Propensity score matching To reduce baseline confounding, we performed propensity score matching (PSM) for each SG subgroup based on clinicodemographic variables including age, race, BMI, hemoglobin A1c(HbA1c), and key comorbidities such as diabetes, hypertension, and chronic kidney disease. Covariates were selected based on their relevance to obesity, indications for bariatric surgery, and cancer risk. Baseline characteristics were assessed within 6 months before the index diagnosis including the day of diagnosis. Sensitivity analysis for unmeasured reproductive factors and HRT use To address potential confounding by unmeasured reproductive factors and hormone replacement therapy (HRT) use, a sensitivity analysis was performed. Patients with relevant reproductive history, high-risk family history indicators, or HRT use recorded anytime prior to the index event (first diagnosis of severe obesity) and within five years after the index event were excluded. Relevant ICD-10 codes included Z34 and Z80.3, and relevant RxNorm codes for HRT medications included 4083, 4099, and 6691. The 5-year incidence of breast cancer, gynecologic cancer, and metastasis was then re-analyzed in this refined cohort. Weight change trajectory analysis Changes in body weight over time were analyzed on the basis of the mean BMI at baseline. Changes in BMI values were assessed within 6 months, 6 to 12 months, 1 to 2 years, 2 to 3 years, 3 to 4 years, and 4 to 5 years after the first diagnosis of severe obesity. This approach allowed for visual representation of weight loss patterns across time points. A two-sided z-test p value of < 0.05 indicated statistical significance. Statistical analysis Both SG and control groups were subjected to 5-year outcome analysis. After PSM, the 5-year cumulative incidence of BC was evaluated. The number of patients with BC, gynecologic cancer (GC) and metastasis in the outcome period were counted for each group. Risk ratios [RRs] were then calculated for each outcome, and the absolute risks compared with two-sided z tests (α = .05). Kaplan-Meier survival analysis was performed to estimate survival probabilities. The incidence of all-cause mortality was identified from deidentified death records within the TriNetX database. BC incidence was identified from relevant ICD-10 codes (eg, C50). GC incidence was identified from relevant ICD-10 codes (eg, C51 to C58 for vulva, vagina, cervix, uterus, ovary, and other female genital organs). The incidence of metastasis was identified from the ICD-10 codes for secondary malignancies of various sites (eg, C77 to C79 and C7B). Results Cohort characteristics The early intervention, intermediate intervention, and delayed intervention subgroups comprised 20321, 13349, and 15276 patients, respectively (Fig. 1A and B). The baseline characteristics of the study cohort are presented in Supplementary Tables 1 to 3. Baseline differences were observed between the SG and control groups before propensity score matching (PSM), with the SG group being younger and having distinct racial and clinical profiles. After 1:1 PSM, covariates including age, race, comorbidities, HbA1c levels and BMI were well balanced across groups, with only negligible residual differences. The final matched cohorts were used to assess 5-year outcomes, including breast cancer, gynecologic cancer, metastasis, and all-cause mortality (Supplementary Tables 1–3). A flowchart depicting patient selection and group allocation is presented in Fig. 1B. Results of weight change trajectory analysis Comparing longitudinal trends in BMI across different SG timings (Fig. 2) revealed distinct weight change trajectories among the SG subgroups. The early intervention subgroup exhibited rapid weight loss within the first 24 months and then maintained a BMI of approximately 33 to 35 kg/m2 (Fig. 2A and B). The intermediate intervention subgroup exhibited gradual weight gain within the first 6 months, followed by gradual weight loss between 6 and 24 months, with BMI stabilizing at approximately 35 kg/m2 (Fig. 2A–C). The delayed intervention subgroup exhibited a slight weight gain within the first 12 months, followed by weight reduction between 12 and 48 months, with BMI stabilizing at approximately 35 kg/m2. By contrast, the control group exhibited a slight reduction in BMI within the first 6 months but still had obesity (mean BMI >40 kg/m2). These findings highlight the persistent challenge of nonsurgical weight management in this patient population and underscore that earlier SG generally leads to more substantial and sustained weight loss.Fig. 2Trajectories of Body Weight Changes in the SG and Control Groups.(A) Distribution of BMI (kg/m2) over a 5-year period in the SG (purple) and control (blue) groups. The SG group was divided as follows on the basis of surgery timing: early intervention (left panel), intermediate intervention (middle panel), and delayed intervention (right panel) subgroups. Each histogram represents a specific follow-up time point (index date, 1 day to 6 months, 6 to 12 months, 1 to 2 years, 2 to 3 years, 3 to 4 years, and 4 to 5 years). The mean BMI values of the SG and control groups are presented in the upper right corner of each panel. (B to D) Line plots depicting the mean BMI trajectories over a period of 60 months for the control group and the (B) early intervention, (C) intermediate intervention, and (D) delayed intervention subgroups. Abbreviation: BMI, body mass index; SG, sleeve gastrectomy. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Trajectories of Body Weight Changes in the SG and Control Groups. (A) Distribution of BMI (kg/m2) over a 5-year period in the SG (purple) and control (blue) groups. The SG group was divided as follows on the basis of surgery timing: early intervention (left panel), intermediate intervention (middle panel), and delayed intervention (right panel) subgroups. Each histogram represents a specific follow-up time point (index date, 1 day to 6 months, 6 to 12 months, 1 to 2 years, 2 to 3 years, 3 to 4 years, and 4 to 5 years). The mean BMI values of the SG and control groups are presented in the upper right corner of each panel. (B to D) Line plots depicting the mean BMI trajectories over a period of 60 months for the control group and the (B) early intervention, (C) intermediate intervention, and (D) delayed intervention subgroups. Abbreviation: BMI, body mass index; SG, sleeve gastrectomy. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Effect of SG timing on the 5-year incidence of all-cause mortality Obesity is closely associated with various metabolic syndrome-related comorbidities, such as type 2 diabetes, hypertension, cardiovascular disease, and certain cancers, all of which contribute to an increased risk of mortality. However, whether surgery timing is associated with long-term survival remains unclear. Our findings revealed that the 5-year incidence of all-cause mortality was lower in all SG subgroups than in the control group (Fig. 3A–C). The early intervention subgroup was most strongly associated with a lower risk (HR: 0.473; 95% CI: 0.381 to 0.588; P < .0001), followed by the intermediate intervention subgroup (HR: 0.518; 95% CI: 0.415 to 0.647; P < .0001) and the delayed intervention subgroup (HR: 0.369; 95% CI: 0.300 to 0.455; P < .0001).Fig. 3Effect of SG Timing on 5-Year Incidence of All-Cause Mortality in Women With Severe Obesity.Kaplan-Meier survival analyses comparing 5-year all-cause mortality between the SG and control groups. (A) The 5-year incidence of all-cause mortality was significantly higher in the early intervention subgroup (blue line) than in the control group (red line). (B) It was significantly higher in the intermediate intervention subgroup (blue line) than in the control group (red line). (C) It was significantly higher in the delayed intervention subgroup (blue line) than in the control group (red line). (D and E) Direct comparisons of the SG subgroups revealed no significant differences in survival between the (D) early intervention (red line) and intermediate intervention (blue line) subgroups or (E) between the early intervention (red line) and delayed intervention (blue line) subgroups. Abbreviation: SG, sleeve gastrectomy. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Effect of SG Timing on 5-Year Incidence of All-Cause Mortality in Women With Severe Obesity. Kaplan-Meier survival analyses comparing 5-year all-cause mortality between the SG and control groups. (A) The 5-year incidence of all-cause mortality was significantly higher in the early intervention subgroup (blue line) than in the control group (red line). (B) It was significantly higher in the intermediate intervention subgroup (blue line) than in the control group (red line). (C) It was significantly higher in the delayed intervention subgroup (blue line) than in the control group (red line). (D and E) Direct comparisons of the SG subgroups revealed no significant differences in survival between the (D) early intervention (red line) and intermediate intervention (blue line) subgroups or (E) between the early intervention (red line) and delayed intervention (blue line) subgroups. Abbreviation: SG, sleeve gastrectomy. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) Intergroup comparisons of baseline characteristics before and after PSM are presented in Supplementary Tables 4 and 5 No significant difference was observed in overall survival between the early and intermediate intervention subgroups (HR: 0.994; 95% CI: 0.749 to 1.319; P = .967; Fig. 3D) or between the early and delayed intervention subgroups (HR: 1.104; 95% CI: 0.836 to 1.458; P = .487; Fig. 3E). Taken together, SG shows a significant association with reduced all-cause mortality in patients with severe obesity. However, SG timing does not appear to have a clear association with survival benefit. Association between SG timing and 5-year incidence of cancer and metastasis To further investigate the long-term oncologic implications of bariatric surgery, we investigated whether SG timing associates with the 5-year incidence of BC, GC, and metastasis in patients with severe obesity. Compared with the control group, the early intervention subgroup was associated with a significantly low incidence of BC (RR: 0.799; 95% CI: 0.647 to 0.987; P = .037), GC (RR: 0.490; 95% CI: 0.378 to 0.635; P < .001), and metastasis (RR: 0.306; 95% CI: 0.230 to 0.408; P < .001; Supplementary Table 6; Fig. 4A). These findings suggest that early bariatric surgery may be associated with a lower risk of cancer development and progression. Both intermediate and delayed SG were associated with a reduced risk of metastasis [intermediate intervention subgroup, RR: 0.374 (95% CI: 0.262 to 0.532; P < .001)]; [delayed intervention subgroup, RR: 0.552 (95% CI: 0.414 to 0.735; P < .001)]; Supplementary Table 6; Fig. 4B and C). However, a delay in SG diminished its protective effect against cancer incidence. Compared with the control group, the intermediate intervention subgroup exhibited no reduction in BC incidence (RR: 0.854; 95% CI: 0.653 to 1.118; P = .25), and the delayed intervention subgroup exhibited a significant increase in BC risk (RR: 1.404; 95% CI: 1.107 to 1.781; P = .005; Supplementary Table 6; Fig. 4C). Moreover, the delayed intervention subgroup exhibited no reduction in the incidence of GC or metastasis. In fact, delayed SG was associated with increased risks of BC, GC, and metastasis (BC, RR: 1.570 [95% CI: 1.210 to 2.038; P = .001]; GC, RR: 2.393 [95% CI: 1.733 to 3.306; P < .001]; metastasis, RR: 1.778 [95% CI: 1.183 to 2.673; P = .005]; Fig. 4D and E). To evaluate the robustness of our findings to potential confounding by reproductive factors and HRT use, we performed a sensitivity analysis excluding patients with relevant reproductive history, high-risk family history, or HRT use. The results were consistent with the main analysis, showing similar trends in the 5-year incidence of breast cancer, gynecologic cancer, and metastasis across SG timing subgroups (Supplementary Table 7). In summary, our findings underscore the importance of timely bariatric surgery in patients with severe obesity. Delaying SG may diminish its protective effect, increasing the risk of cancer.Fig. 4Effect of SG Timing on 5-Year Incidence of Cancer in Women With Severe Obesity.Logistic regression analyses were performed to evaluate the 5-year incidence of BC, GC, and metastasis in the SG and control groups. (A to C) Forest plots depicting the effects of (A) early, (B) intermediate, and (C) delayed SG on the incidence of BC, GC, and metastasis. (D and E) Effects of (D) intermediate and (E) delayed SG on the incidence of BC, GC, and metastasis. Abbreviations: BC, breast cancer; GC, gynecologic cancer; SG, sleeve gastrectomy. Effect of SG Timing on 5-Year Incidence of Cancer in Women With Severe Obesity. Logistic regression analyses were performed to evaluate the 5-year incidence of BC, GC, and metastasis in the SG and control groups. (A to C) Forest plots depicting the effects of (A) early, (B) intermediate, and (C) delayed SG on the incidence of BC, GC, and metastasis. (D and E) Effects of (D) intermediate and (E) delayed SG on the incidence of BC, GC, and metastasis. Abbreviations: BC, breast cancer; GC, gynecologic cancer; SG, sleeve gastrectomy. Results of age-stratified analysis To investigate how age may influence the association between SG timing and cancer incidence, we performed an age-stratified analysis, focusing on BC and GC. In younger patients (age <45 years), neither early nor delayed SG significant reduction of the 5-year incidence of BC was observed (early intervention subgroup, RR: 1.119 [95% CI: 0.760–1.645; P = .569]; intermediate intervention subgroup, RR: 1.103 [95% CI: 0.661 to 1.841; P = .707]; delayed intervention subgroup, RR: 1.051 [95% CI: 0.674 to 1.640; P = .826]; Fig. 5A). This finding suggests that SG does not confer protection against BC in younger women with obesity. By contrast, in older women (age ≥45 years), SG timing played a key role in determining 5-year outcomes. Early SG significant reduction of the risk of BC was observed (RR: 0.649; 95% CI: 0.512 to 0.822; P < .0001). In the intermediate intervention subgroup, the association was not statistically significant (RR: 0.808; 95% CI: 0.594 to 1.098; P = .172). In the delayed intervention subgroup, a significant increase in was observed in BC risk (RR: 1.539; 95% CI: 1.171 to 2.024; P = .002; Fig. 5A).Fig. 5Effects of SG Timing on Cancer Incidence in Younger and Older Women With Severe Obesity.Logistic regression analyses were performed to evaluate the 5-year incidence of (A) breast cancer and (B) gynecologic cancer in the SG subgroups and control group. Each analysis was further stratified by age, with younger patients (age <45 years) and older patients (age ≥45 years) analyzed separately to evaluate the age-specific effects of SG timing on cancer risk. Abbreviation: SG, sleeve gastrectomy. Effects of SG Timing on Cancer Incidence in Younger and Older Women With Severe Obesity. Logistic regression analyses were performed to evaluate the 5-year incidence of (A) breast cancer and (B) gynecologic cancer in the SG subgroups and control group. Each analysis was further stratified by age, with younger patients (age <45 years) and older patients (age ≥45 years) analyzed separately to evaluate the age-specific effects of SG timing on cancer risk. Abbreviation: SG, sleeve gastrectomy. The results presented in Fig. 5B align with those presented in Fig. 4, which indicate that SG timing contributes to reduce the incidence of GC in both younger and older patients with severe obesity. Early SG was significantly associated with a reduced incidence of GC in both age groups (younger women, RR: 0.557 [95% CI: 0.350 to 0.885; P = .012]; older women, RR: 0.510 [95% CI: 0.377 to 0.696; P < .001]). Delaying SG was associated with a weaker association with cancer risk reduction younger women, RR: 0.550 [95% CI: 0.322 to 0.937; P = .026]; older women, RR: 0.749 [95% CI: 0.530 to 1.059; P = .101]). In the delayed intervention subgroup, the risk of GC was significantly higher in younger patients than in older patients (RR: 1.828; 95% CI: 1.161 to 2.877; P = .008). These findings highlight the importance of early intervention, particularly in older patients, in whom timely surgery significantly reduces the risk of BC. Regarding GC, early SG benefits both younger and older patients, but delaying surgery can negate these benefits and may even increase cancer risk. Therefore, clinicians should consider both patient age and surgery timing when devising strategies for optimizing the cancer-preventive effects of bariatric surgery. Discussion We analyzed data from TriNetX, a large-scale electronic health record network, to investigate the association between SG timing and 5-year incidence of BC, GC, metastasis, and all-cause mortality in women with severe obesity. To the best of our knowledge, this study is the first to reveal a complex, time-dependent association between SG and BC risk while highlighting its potential associations with weight loss and cancer prevention in different age groups undergoing surgery at different time points. Studies have revealed a strong, positive association between obesity and BC risk [[3], [4], [5]]. Bariatric surgery, including SG, has been associated with reduced risks of cancer (eg, BC) and cancer-related mortality [[10], [11], [12],[17], [18], [19], [20], [21]]. These findings reinforce the overall health benefits of weight loss in individuals with obesity. The mechanisms underlying obesity-driven carcinogenesis are multifactorial and include chronic low-grade inflammation, altered adipokine profiles (eg, elevated leptin and reduced adiponectin levels), insulin resistance and hyperinsulinemia, and dysregulated sex hormone metabolism (eg, increased aromatization of androgens to estrogens in adipose tissue), all of which contribute to a protumorigenic microenvironment [22,23]. SG, by inducing substantial and sustained weight loss, effectively reverses many of the aforementioned metabolic and hormonal disturbances [24]. The protective effects of bariatric surgery are therefore attributable to substantial reductions in body weight, improvements in metabolic parameters, alleviation of chronic inflammation, and normalization of sex hormone levels [25,26]. Although the general protective effects of bariatric surgery are well known, very few studies have focused specifically on surgery timing, particularly the preoperative duration of obesity and its correlation with BC incidence. In the present study, we addressed this research gap by stratifying women with severe obesity by surgery timing and comparing various 5-year outcomes among the subgroups. The most notable and unexpected finding of our study pertains to the association between SG timing and BC incidence (Supplementary Table 6; Fig. 4). The weight loss trajectories depicted in Fig. 2 corroborate the importance of SG timing. Earlier SG was associated with more rapid, pronounced, and sustained weight loss. Our study reported interesting findings that early intervention subgroup exhibited a significant reduction in BC risk. However, this protective effect diminished and became nonsignificant in the intermediate intervention subgroup. Notably, the delayed intervention subgroup exhibited a significant increase in BC risk. This unexpected reversal in cancer risk is a major finding of our study. This paradoxical increase in BC risk in the delayed intervention subgroup should be interpreted with caution. As this is an observational study based on electronic health records, detection bias cannot be fully excluded. Women who delayed surgery may have had more frequent healthcare encounters, which could increase the likelihood of cancer detection rather than reflect a true increase in biological incidence. Furthermore, the observed association may be partially attributable to residual confounding by unmeasured variables, such as disparities in healthcare utilization patterns, differential screening intensity, or the existence of pre-existing yet undiagnosed subclinical malignancies. From a biological perspective, prolonged exposure to severe obesity prior to surgical intervention may result in sustained systemic inflammation, hyperinsulinemia, insulin resistance, and increased estrogen production through adipose tissue aromatase activity. These obesity-related mechanisms have been implicated in breast carcinogenesis and may partially explain the elevated risk observed in the delayed intervention subgroup. Therefore, our findings should be interpreted as an association rather than evidence of a direct causal effect of delayed surgery on breast cancer risk. Prolonged obesity leads to chronic inflammation, sustained hyperinsulinemia, and long-term alterations in sex hormone profiles, thereby contributing to a protumorigenic microenvironment. Surgical intervention cannot fully mitigate this microenvironment; in some cases, surgery may even unmask it by altering baseline conditions [27]. “Obesity memory” suggests that prolonged exposure to obesity-related metabolic and inflammatory insults can induce epigenetic changes and cellular alterations that persist even after weight loss, predisposing tissue to disease [28,29]. In the context of BC, prolonged obesity may lead to irreversible changes in key signaling cascades or immune cell populations within the mammary gland, fostering an environment conducive to cancer development and progression [30]. Obesity drives distinct molecular changes in breast tissue, promoting cellular proliferation, increasing angiogenesis, and altering immune responses, which may become ingrained over time and hinder risk reversal, even after weight loss [[31], [32], [33]]. The fact that our delayed intervention subgroup exhibited a reduction in BMI after SG but still had an increased risk of BC indicates that the duration of obesity or its long-term biological consequences, rather than BMI at the time of surgery, determines the risk of cancer. In this study, “obesity diagnosis” was defined as either a BMI ≥35 kg/m2 or the first recorded ICD-10 code for obesity (E66.0–E66.9) in the electronic health record. Importantly, our findings should not be interpreted as a rigid mandate that surgery must be performed within 6 months. In real-world practice, we recognize that such a timeline is often difficult to achieve due to insurance requirements, preoperative optimization, and patient decision-making processes. Rather, the observed association highlights the potential importance of avoiding prolonged delays when clinically feasible. As clinicians involved in the care of patients with both obesity and breast cancer risk, these findings have influenced how we frame discussions with patients. We are now more attentive to the potential impact of treatment timing and more proactive in facilitating weight management and timely referral for metabolic surgery when appropriate. In this context, the <6-month interval should be viewed not as a strict requirement, but as a clinically meaningful temporal reference point that may help guide shared decision-making and prioritization of care. SG exerted no significant association with the incidence of BC in younger women with severe obesity. This finding is consistent with that of a meta-analysis indicating that bariatric surgery exerts no significant effect on the risk of early-onset colon cancer [17]. The age-based differences we observed in the association between SG and BC incidence may reflect distinct mechanisms underlying tumor development across age groups. In younger patients, the risk of BC is frequently influenced by hereditary factors, particularly germline BRCA1/BRCA2 mutations, rather than by obesity-driven metabolic disturbances [34]. Although obesity may still exacerbate disease progression in younger patients, the primary etiologic drivers appear to be less amenable to metabolic intervention [35,36]. Therefore, SG may not effectively target the dominant pathways of cancer initiation in this population. In contrast, the more pronounced association observed in women aged ≥45 years suggests that metabolic and hormonal mechanisms may play a greater role in tumorigenesis in this population. In peri- and postmenopausal women, adipose tissue becomes a major source of estrogen production through aromatization, and obesity-related hyperinsulinemia and chronic inflammation may further promote tumor initiation and progression [22,23,30]. In this context, earlier metabolic intervention may have greater potential to modify these pathways and reduce long-term cancer risk. From a clinical perspective, these findings may have implications for surgical prioritization. For women with severe obesity approaching or exceeding midlife, avoiding unnecessary delays in metabolic surgery, when feasible, may play a critical role in oncologic risk reduction. Conclusion This study suggests that SG not only significantly correlated with reduction of the 5-year incidence of all-cause mortality in patients with severe obesity but also confers substantial protection against BC and GC. Notably, SG timing was significantly associated with differences in outcomes, with early SG showing the strongest associations with improved survival and lower cancer incidence. Furthermore, older patients demonstrated a notable association between SG and survival outcomes. Taken together, these findings underscore the importance of timely bariatric surgery and support the consideration of SG as a proactive strategy for mitigating cancer and mortality risks. Ethical approval The study protocol was approved by the Institutional Review Board of Taipei Tzu Chi Hospital (approval number: 14-IRB015). Consent for publication Not applicable. Ethics committee approval and patient consent This study was approved by the Institutional Review Board of Taipei Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Taiwan (Approval Number: 14-IRB015). Consent for publication All authors have read and approved the final manuscript. Funding This work was supported by 10.13039/100020595National Science and Technology Council (113-2314-B-303-014 and 114-2314-B-016), 10.13039/501100008108Taipei Tzu Chi Hospital and 10.13039/501100005925Buddhist Tzu Chi Medical Foundation (TCRD-TPE-MOST-111-15 and TCMF-CM3-112-03) and through collaboration between Tzu Chi Hospital and 10.13039/501100001869Academia Sinica (TCAS-112-02 and TCAS-115-01). CRediT authorship contribution statement Kian-Hwee Chong: Investigation, Writing – original draft. Yao-Jen Chang: Writing – review & editing. Joshua Wang: Data curation, Formal analysis, Methodology. Kuo-Cheng Lu: Supervision, Writing – review & editing. Tsai-Chieh Hou: Data curation, Validation. Ching-Feng Cheng: Conceptualization, Investigation, Writing – review & editing. Kuo-Wang Tsai: Supervision. Declaration of competing interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.