肥胖患者减重手术与痴呆及帕金森病风险:一项 TriNetX 纵向队列研究
Bariatric surgery and risk of dementia and Parkinson’s disease in patients with obesity
基于 TriNetX 网络的纵向队列研究显示,减重手术与肥胖患者痴呆(HR 0.48,95%CI 0.39–0.58)和帕金森病(HR 0.70,95%CI 0.51–0.96)发生率降低相关。匹配后两组各 32552 人,同时观察到心血管疾病、肥胖相关癌症和全因死亡发生率较低。排除术后 6 个月内结局后帕金森病关联不再有统计学证据,且观察性设计不能推定因果关系。
Abstract
Objective
We aimed to evaluate the association between bariatric surgery (BS) and the incidence of dementia and Parkinson’s disease. To assess consistency with established benefits of BS, we examined its association with cardiovascular disease, obesity-associated cancers, and all-cause mortality.
Methods
We conducted a longitudinal cohort study using electronic health records from TriNetX network. Patients aged ≥16 years with at least one BMI record of ≥35 kg/m2 between 1st Jan 2005 and 31st Dec 2022 were included. The exposed cohort included those who underwent a BS; the unexposed cohort included everyone without BS. The index date was the earliest date for BS up to 6 months after a BMI of ≥35 kg/m2 for the exposed, and the earliest encounter code up to 6 months after a BMI of ≥35 kg/m2 for the unexposed. Individuals were 1:1 matched using propensity score matching on covariates age at index date, gender, ethnicity, race, BMI, hypertension, type 2 diabetes (T2D), socioeconomic status, smoking, alcohol, and history of neurological diseases. Cox proportional hazards regressions were performed for each outcome.
Results
After matching, there were 32,552 individuals each in the BS and non-BS group. The median follow-up was 4.7 years in individuals with BS and 5.9 years in those without. Compared to the non-BS group, individuals who underwent BS had significantly reduced incidence rates of dementia (hazard ratio 0.48, 95% CI: 0.39–0.58) and Parkinson’s disease (0.70; 0.51–0.96). Similar associations were observed for CVD (0.53; 0.51–0.56); obesity-associated cancers (0.66; 0.60–0.72) and all-cause death (0.46; 0.41–0.51). Survival probability at 20-years was 84.7% (95% CI: 71.1–92.3%) in the BS group vs 81.8% (77.5–85.4%) in those without.
Conclusion
BS was associated with a lower incidence of dementia and Parkinson’s disease in individuals with BMI ≥ 35 kg/m2, in addition to reducing CVD, obesity-associated cancers, and mortality, suggesting potential neuroprotective and systemic health benefits.
Introduction
Obesity is a major global health challenge and a leading cause of morbidity and premature mortality. It is strongly associated with a range of diseases including cardiovascular disease, type 2 diabetes (T2D), and several obesity-associated cancers [1, 2]. In addition to these well-established associations, there is growing recognition of obesity as a potential risk factor for neurodegenerative conditions such as dementia and Parkinson’s disease [3,4,5]. This link may be mediated through obesity-related metabolic dysfunction, including insulin resistance, T2D and hypertension, as well as through biological mechanisms such as chronic systemic inflammation, oxidative stress, vascular dysfunction and altered lipid metabolism, all of which can lead to neuronal damage and cognitive decline [6, 7]. These pathophysiological pathways suggest that obesity may play a role in the onset and progression of dementia and Parkinson’s disease.
Despite the advances on pharmacological therapies for obesity and T2D [such as the newer generation of Glucagon-Like Peptide-1 (GLP-1) receptor analogue (RA) - based treatments], bariatric surgery (BS) remains the most effective and cost-effective intervention achieving substantial and sustained long-term weight loss (WL) in people with severe obesity [8,9,10]. The recently updated criteria from the International Federation for the Surgery of Obesity and Metabolic Disorders (IFSO) recommend BS for individuals with a body mass index (BMI) ≥ 35 kg/m² (regardless of obesity-related complications), or for those with BMI ≥ 30 kg/m² who have metabolic disease and have not achieved adequate WL or metabolic control through non-surgical interventions [11]. The two most commonly performed bariatric procedures worldwide [sleeve gastrectomy (SG) and Roux-en-Y gastric bypass (RYGB)] typically result in 25–33% total body WL in the first two postoperative years and sustained 20–25% WL over time [12, 13]. This effect is thought to be largely attributed to reduced food intake, likely as a consequence of decreased appetite. In procedures such as RYGB and SG, reduced appetite appears to be predominantly mediated by altered postprandial entero-pancreatic hormone responses, characterised by increased GLP-1 and peptide YY (PYY) secretion, together with reduced ghrelin levels, the latter being particularly evident following SG. These changes arise from altered nutrient delivery to the distal intestine and changes in gastric anatomy and gastric emptying [14, 15]. Nevertheless, certain bariatric procedures that involve anatomical re-arrangement of the gastrointestinal tract such as biliopancreatic diversion and RYGB, may also lead to some additional WL through reduced nutrient absorption [16].
BS is also associated with improvement in multiple cardio-metabolic risk factors and reduced incidence of obesity-related complications including T2D, hypertension, cardiovascular disease (CVD), cancer and mortality [17,18,19,20]. However, while the benefits of BS in cardiometabolic health are well established, there is comparatively limited data on its potential role in reducing the risk of developing neurodegenerative diseases, such as dementia and Parkinson’s disease. Some studies suggest that BS may help prevent or improve cognitive decline and reduce the incidence of Alzheimer’s disease and other dementias [21, 22]. Yet, other studies have found an increased incidence of dementia following BS [23]. As for Parkinson’s disease, although previous limited research has suggested a possible link between obesity and the risk of developing Parkinson’s [24, 25], to the best of our knowledge there are no studies examining the association between BS and the incidence of Parkinson’s disease.
In this study, we examined the association between BS with the incidence of dementia and Parkinson’s disease as primary outcomes, using real-world data from the TriNetX network. To assess whether these associations occurred in parallel with the established systemic benefits of BS, we also examined its relationship with cardiovascular disease, obesity-associated cancers (OACs), and all-cause mortality.
Methods
This is a retrospective observational longitudinal cohort study using electronic healthcare records (EHR) data from the TriNetX (TNX) network. The TNX network includes anonymised inpatient and outpatient data on diagnoses, procedures, medications, and laboratory findings recorded in the EHR of participating healthcare organisations (HCOs). On the TNX platform, after defining the cohorts to be compared, only aggregated counts and figures and preset analytical options are accessible, while the underlying individual-level data are not accessible in compliance with US laws governing the security and privacy of healthcare data. We defined our cohorts using data from the US collaboration network, which included data for over 103 million patients in the USA from 70 HCOs.
All TNX curated codes (ICD-9, ICD-10, Current Procedural Terminology) used to define the cohorts, exposure, confounders, and outcomes are available in the Supplementary file 1. We followed the RECORD guidelines [26] in conducting and reporting the results of this study (Supplementary file 8).
Study population and exposure
For all outcomes, all patients who had at least one record of a BMI ≥ 35 kg/m2 up to 6 months before a health service encounter code at ages 16 years or over during the study period from 1st January 2005 to 31st December 2022 were included in the target population. The exposed cohort included those who had a BS event recorded up to 6 months after a BMI record. The date of the earliest BS code recorded up to 6 months after a BMI record of ≥35 kg/m2 during the study period was the index event date for the exposed cohort. The unexposed cohort excluded those who had a record of BS at any time. The date of the earliest encounter code recorded up to 6 months after a BMI record of ≥35 kg/m2 during the study period was the index event date for the unexposed cohort. Patients were included if they had a hospital visit 6 months or more before, or after the index date (codelist in Supplementary file 2).
Patients were excluded from these cohorts if there was a record of dementia, Parkinson’s disease, cardiovascular disease, OACs, or death on or before the index date. A flowchart of steps of cohort selection used on the TNX platform is provided in Fig. 1.
N the number of patients in the TriNetX (TNX) US collaboration network at each step.
Confounding variables
The covariates age at index event (continuous), gender (male, female, unknown), ethnicity (Not Hispanic or Latino, Hispanic or Latino, Unknown), race (White, Black or African American, Asian, American Indian or Alaskan native, native Hawaiian or other Pacific islander, Other, Unknown), BMI (35–40 kg/m2, 40–45 kg/m2, 45–50 kg/m2, >50 kg/m2), hypertensive disease (present/absent), T2D (present/absent), potential health hazards related to socioeconomic and psychosocial circumstances (present/absent), alcohol related disorders (present/absent), nicotine dependence (present/absent), history of nicotine dependence (present/absent), epilepsy and recurrent seizures (present/absent), injury of head (present/absent), depressive episode (present/absent), schizophrenia (present/absent), family history of epilepsy and other diseases of the nervous system (present/absent), hypercholesterolemia (present/absent), hyperglyceridemia (present/absent), and hyperlipidemia (present/absent) were included a priori for propensity score (PS) matching to account for these confounders. The nearest record of these covariates on or before the index event was used.
Outcomes
The incidence of dementia, and Parkinson’s disease was compared across patients in the exposed and unexposed cohorts. Patient follow-up started from the index date and ended when the outcome of interest occurred, the end of study period (31st December 2022), or last record (which may include death date) in TNX database, whichever was earliest. The incidence of Alzheimer’s disease, and vascular dementia was also compared across patients in the exposed and unexposed cohorts to further explore causes of dementia.
We also compared the incidence of CVD, obesity-associated cancers (OACs), and death across patients in the exposed and unexposed cohorts. Previous work indicates that BS is associated with these outcomes. We compared our findings for these outcomes with prior studies in order to gauge our confidence in the interpretation of estimates for dementia and PD given the potential for bias in studies of real-world data.
Statistical methods
Exposed and unexposed cohorts were 1:1 matched using PS matching without replacement; PS matching was conducted by the TNX platform using the “greedy nearest neighbour matching” with the caliper value fixed at 0.1 times the pooled standard deviation of the propensity score [27]. The a priori confounders specified above were included in the PS matching.
The baseline (prior to or at the index date) characteristics for the exposed and unexposed cohorts, before and after matching, were obtained. A standardised mean difference of 0.1 [28] was set a priori to assess the balancing of confounders included in the propensity score matching.
Cox proportional hazards regression was performed on the propensity score matched cohorts to estimate hazards ratios for each outcome of interest. Survival probability and Kaplan–Meier survival curve for the all-cause death outcome is presented.
Sensitivity analyses
We repeated these analyses after excluding outcomes reported up to 6 months after index date to assess impact of potential detection bias on the results. We also performed a sub-group analysis comparing the incidence of dementia, and Parkinson’s disease across patients who had RYGB to those who had SG. Additional comparisons could not be performed due to lower outcome events for other BS types, and their results were obfuscated by the TNX platform for data protection.
Results
Baseline characteristics
Before matching, 32,642 patients were included in the exposed and 1,525,743 in the unexposed cohort (Supplementary file 3); following propensity score matching, each of the two cohorts comprised 32,552 patients. The baseline characteristics of the cohort are presented in Table 1.
Survival analysis
In the matched cohort, the median follow-up was 4.7 years with a maximum follow-up of 19.7 years in the exposed cohort, and 5.9 years with a maximum follow-up of 20 years in the non-exposed cohort.
Compared to those who did not undergo a BS, those with a BS intervention had 52.4% lower incidence rates of dementia (HR 0.48, 95% CI: 0.39–0.58) and 30.1% lower incidence rates of Parkinson’s disease (HR 0.70, 95% CI: 0.51–0.96) (Fig. 2). Those with a BS intervention had 42.3% lower incidence rates of Alzheimer’s disease (HR 0.58, 95% CI: 0.38–0.88) and 73.1% lower incidence rates of vascular dementia (HR 0.27, 95% CI: 0.15–0.48) compared to those who did not undergo a BS (Supplementary file 4).
Hazard ratios with 95% Confidence intervals (HR 95% CI) comparing the incidence rates of those with a bariatric surgery (BS) to those without, for all outcomes. Legend: CVD cardiovascular disease, OACs obesity-associated cancers.
BS was associated with 46.9% lower incidence rates of CVD (HR 0.53, 95% CI: 0.51–0.56) and 34.3% lower incidence rates of OACs (HR 0.66, 95% CI: 0.60–0.72).
BS was associated with 54.3% lower (HR 0.46, 95% CI: 0.41–0.51) incidence rates of all-cause death compared to those who did not have a BS (Fig. 2); the survival probability at 20-years in those who had a BS was 84.7% (95% CI: 71.1–92.3%) compared to 81.8% (95% CI 77.5–85.4%) in those who did not (Fig. 3).
Kaplan–Meier survival curves for death outcome, comparing those with a bariatric surgery (BS) to those without, with 95% Confidence intervals.
Sensitivity analyses
When patients who had an outcome up to 6 months after index date were excluded, 32,303 patients were included in the exposed and 1,475,594 in the unexposed cohort before matching (Supplementary file 5); following propensity score matching, each of the two cohorts comprised 32,203 patients. Compared to those who did not undergo a BS, those with a BS intervention had 47.5% lower incidence rates of dementia (HR 0.52, 95% CI: 0.42–0.65), while there was no evidence of an association with incidence of Parkinson’s disease (Supplementary file 6).
In comparisons between Roux-en-Y gastric bypass (RYGB) and sleeve gastrectomy (SG), there was no evidence of a difference in the incidence of dementia or Parkinson’s disease (Supplementary file 7).
Discussion
Our study showed that having BS was associated with lower incidence rates of both dementia and Parkinson’s disease, compared to matched individuals who did not undergo BS. These findings contribute important evidence to the limited body of literature on the impact of BS on risk for neurodegenerative diseases.
Our results are consistent with the study by Chen et al. [21], using data from 5,303 BS patients PS matched to 10,606 patients without BS, which reported a lower incidence of dementia following BS (subdistribution HR 0.37; 95% CI: 0.15– 0.89). Additionally, Hathaway et al. [22] reported in their systematic review that cognition appears likely to improve after BS. In contrast, Kim et al. [23], using data from 17,026 BS patients to 34,052 patients without BS, reported an increased risk for dementia (HR 1.33, 95% CI: 1.06–1.67) following BS. This could potentially be due to post-surgical nutritional deficiencies, such as Vitamin B12 and D [29,30,31], which may adversely affect cognitive functions if not managed properly. However, that study relied on Utah BS registries and a general population comparison from the Utah Population Database, raising concerns about selection bias. Furthermore, it did not exclude individuals with pre-existing cardiovascular disease or cancer – conditions that may contraindicate surgery – introducing further bias. Additionally, data from 1996 may suffer from poor exposure and outcome ascertainment.
Our findings also showed that BS was associated with lower incidence rates of Parkinson’s disease, compared to those who did not have BS. This could potentially be linked with increased dopamine receptor availability following BS [32, 33]. Moreover, WL due to BS could be beneficial for brain function by reducing systemic inflammation, improving insulin sensitivity, and enhancing vascular health [34, 35]. Notably, we could not find any prior studies examining the association between WL induced by BS and Parkinson’s disease incidence.
GLP-1 appears to have multiple effects on the central nervous system, including anti-inflammatory activity, improved mitochondrial function and antioxidative stress effects [36, 37]. Bariatric procedures such as RYGB and SG are known to markedly increase postprandial GLP-1 secretion [38], which may, at least in part, contribute to the neuroprotective effects observed in our study. Trials are also investigating GLP-1-based therapies as potential treatments for dementia [39,40,41] and Parkinson’s disease [42, 43]. In studies of dementia, while early-phase studies with liraglutide (1.2–1.8 mg once daily) have suggested potential cognitive and neuroimaging benefits in Alzheimer’s disease [44], results from recent phase 3 trials did not demonstrate that treatment with the GLP-1RA oral semaglutide (up to 14 mg) was superior to placebo in slowing disease progression [45]. Together, these findings support further investigation into whether enhanced GLP-1 levels, alongside other metabolic and hormonal changes, contribute to the benefits of BS, and how these effects compare with those of pharmacological GLP-1–based therapies. In this context, evidence from post-hoc analyses of clinical trials and observational studies suggests that GLP-1RAs, approved for T2D management, may reduce the risk of cognitive decline and dementia [46,47,48].
Evidence for neuroprotective effects of GLP-1 RAs in Parkinson’s disease is more limited and findings are also inconsistent. A 12-month, multicentre, randomized phase 2 trial found that the GLP-1 RA lixisenatide significantly slowed motor symptom progression in early-stage Parkinson’s disease, with benefits persisting for two months after treatment cessation [42]. In contrast, a larger 96-week phase 3 trial of GLP-1 RA exenatide found no significant effect on Parkinson’s disease progression or symptom improvement [43]. It should be noted, however, that both lixisenatide and exenatide have minimal effects on WL and lesser anti-inflammatory and antioxidative stress effects compared to novel GLP-1-based pharmacotherapies, such as semaglutide and tirzepatide (a dual GIP and GLP-1 receptor agonist) [49]. These mixed findings also underscore the need for further research to clarify whether specific GLP-1RAs provide neuroprotective benefits in PD and to explore the potential for newer, more potent weight-loss agents to confer similar benefits.
Our study confirmed also that BS was associated with lower incidence of cardiovascular disease (CVD), OACs, and all-cause mortality compared to those who did not have BS. Our results on CVD disease are consistent with previous research reporting 50–54% lower incidence of heart failure, 37–42% lower incidence of myocardial infarction, 29–36% lower incidence of stroke, and 70% lower incidence of major adverse cardiovascular events (MACE) following BS [50,51,52], including studies using TNX data [53,54,55]. Previous studies have also reported a 28–44% lower incidence of cancer overall following BS, with reduction of 41% for OACs such as breast, endometrial, liver and colorectal [56,57,58,59]. Our study results also suggested a lower incidence of cancer in people who underwent BS, though the observed effect size was smaller. Additionally, studies have reported that individuals who had a BS had around 41–74.6% lower incidence of all-cause death [20, 52, 60]. Our study results were broadly similar to these studies as well as with previously published studies using TNX data [54, 61]. This alignment with previous studies enhances confidence and reinforces the reliability of the observed associations between dementia and PD.
This study has significant strengths. We had access to EHR clinical and demographic data, with long follow-up from a large population in the USA. This enabled us to define well-matched cohorts of individuals who underwent BS and those who did not. The large sample size provided sufficient numbers to estimate associations with good precision and power. The inclusion of both emerging obesity-related outcomes, such as dementia and Parkinson’s disease, and multiple traditional outcomes provide a more holistic view of the long-term impact of BS.
Our study has also important limitations due to its observational study design. Selection bias is possible as only individuals with a health service encounter up to 6 months after a BMI record ≥35 kg/m² were included. Outcome-detection bias is possible if individuals who had BS had more contact with the health system. Calendar time bias, and longer observation time for the unexposed are possible as the index date for the unexposed cohort was the earliest encounter date up to 6 months after BMI record of at least 35 kg/m2 while for the exposed cohort it was the date of the earliest BS event up to 6 months after BMI record of at least 35 kg/m2. As the underlying data was not accessible, index dates were defined independently for each cohort and could not be defined based on the matched exposed participants and may therefore have introduced immortal time bias, potentially affecting the strength of observed associations. Due to the observational nature of this study, we cannot attribute causality to our results. These results might be affected by unmeasured confounding e.g. due to education level and insurance coverage information not available in TNX; and misclassification bias, such as misclassified exposure timing, or outcome coding bias e.g. due to underdiagnosis, miscoding, or delayed recording of dementia and Parkinson’s disease. The TNX database only includes inpatient and outpatient data from EHRs, and we cannot be certain of complete follow-up of all patients. We used diagnoses recorded in TNX as proxy indicators for several potential confounders, such as socioeconomic status, smoking, alcohol use, and lipid levels. However, these proxy indicators might not fully capture differences in socioeconomic circumstances, lifestyle behaviours, or healthcare access, and therefore, residual confounding related to these factors remains possible. We did not adjust for baseline diabetes medication such as metformin and GLP 1RAs, as these medications may be initiated, discontinued, or modified as a consequence of BS and resulting changes in metabolic health. Consequently, these medications may lie on the causal pathway between BS and the outcomes of interest, and therefore should not be considered as confounders and adjusted for. Individuals who undergo BS may have greater access to healthcare, including more comprehensive health insurance coverage, which could be associated with better health outcomes independent of the surgical intervention. Consequently, some of the lower incidence rates observed in the BS cohort may reflect differences in underlying healthcare access rather than the effects of BS alone [62]. In addition, individuals undergoing BS typically receive extensive pre-operative evaluation and treatment of obesity-related conditions, which may not apply to unexposed individuals, potentially introducing bias. There may also be confounding by indication, as individuals in the unexposed group may have been ineligible for BS. We used an aggregated codelist encompassing all types of BS, with the earliest BS used as the index event. The number of Parkinson’s disease events was relatively low, potentially limiting statistical power and increasing the likelihood of chance findings. Finally, these results may not be generalisable to populations with different sociodemographic characteristics or healthcare systems than those in the US.
Conclusion
In summary, this study found that BS was associated with lower incidence of dementia and Parkinson’s disease in individuals with severe obesity. These results extend the known benefits of surgical WL into the domain of neurodegenerative disease prevention and align with prior research on the cardiometabolic and oncologic benefits of BS. Further research is needed to address the limitations of this analysis.
Data availability
TriNetX data used for current study were used under license and are not publicly available.
References
Di Angelantonio E, Bhupathiraju SN, Wormser D, Gao P, Kaptoge S, de Gonzalez AB, et al. Body-mass index and all-cause mortality: individual-participant-data meta-analysis of 239 prospective studies in four continents. Lancet. 2016;388:776–86.
Lingvay I, Cohen RV, CWl Roux, Sumithran P. Obesity in adults. Lancet. 2024;404:972–87.
Chen S, Kuper H. Tracing the temporal trends of modifiable risk factors in dementia: insights from the English Longitudinal Study of Ageing (2004–2019). Lancet. 2023;402:S34.
Chen J, Guan Z, Wang L, Song G, Ma B, Wang Y. Meta-analysis: overweight, obesity, and Parkinson′s disease. Int J Endocrinol. 2014;2014:203930.
Tang X, Zhao W, Lu M, Zhang X, Zhang P, Xin Z, et al. Relationship between central obesity and the incidence of cognitive impairment and dementia from cohort studies involving 5,060,687 participants. Neurosci Biobehav Rev. 2021;130:301–13.
Ghowsi M, Qalekhani F, Farzaei MH, Mahmudii F, Yousofvand N, Joshi T. Inflammation, oxidative stress, insulin resistance, and hypertension as mediators for adverse effects of obesity on the brain: a review. Biomedicine. 2021;11:13–22.
Rojas-Gutierrez E, Muñoz-Arenas G, Treviño S, Espinosa B, Chavez R, Rojas K, et al. Alzheimer’s disease and metabolic syndrome: a link from oxidative stress and inflammation to neurodegeneration. Synapse. 2017;71:e21990.
Essop T, Tran K, Purdy AC, Daly SC. Comparative effects of GLP-1 agonists, sleeve gastrectomy and Roux-en-Y gastric bypass on diabetes mellitus outcomes. Curr Diab Rep. 2024;24:273–89.
Picot J, Jones J, Colquitt JL, Gospodarevskaya E, Loveman E, Baxter L, et al. The clinical effectiveness and cost-effectiveness of bariatric (weight loss) surgery for obesity: a systematic review and economic evaluation. Health Technol Assess. 2009;13:1–iv.
Docimo S Jr., Shah J, Warren G, Ganam S, Sujka J, DuCoin C. A cost comparison of GLP-1 receptor agonists and bariatric surgery: what is the break even point? Surg Endosc. 2024;38:6560–5.
Eisenberg D, Shikora SA, Aarts E, Aminian A, Angrisani L, Cohen RV, et al. 2022 American Society for Metabolic and Bariatric Surgery (ASMBS) and International Federation for the Surgery of Obesity and Metabolic Disorders (IFSO): indications for metabolic and bariatric surgery. Surg Obes Relat Dis. 2022;18:1345–56.
O’Brien PE, Hindle A, Brennan L, Skinner S, Burton P, Smith A, et al. Long-term outcomes after bariatric surgery: a systematic review and meta-analysis of weight loss at 10 or more years for all bariatric procedures and a single-centre review of 20-year outcomes after adjustable gastric banding. Obes Surg. 2019;29:3–14.
Adams TD, Davidson LE, Litwin SE, Kim J, Kolotkin RL, Nanjee MN, et al. Weight and metabolic outcomes 12 years after gastric bypass. N Engl J Med. 2017;377:1143–55.
Casimiro I, Sam S, Brady MJ. Endocrine implications of bariatric surgery: a review on the intersection between incretins, bone, and sex hormones. Physiol Rep. 2019;7:e14111.
Papamargaritis D, le Roux CW. Do gut hormones contribute to weight loss and glycaemic outcomes after bariatric surgery? Nutrients. 2021;13:762.
Al-Najim W, Docherty NG, le Roux CW. Food intake and eating behavior after bariatric surgery. Physiol Rev. 2018;98:1113–41.
Buchwald H, Avidor Y, Braunwald E, Jensen MD, Pories W, Fahrbach K, et al. Bariatric surgery a systematic review and meta-analysis. JAMA. 2004;292:1724–37.
Sjöström L, Narbro K, Sjöström CD, Karason K, Larsson B, Wedel H, et al. Effects of bariatric surgery on mortality in Swedish Obese Subjects. N Engl J Med. 2007;357:741–52.
Sjöström L. Review of the key results from the Swedish Obese Subjects (SOS) trial - a prospective controlled intervention study of bariatric surgery. J Intern Med. 2013;273:219–34.
Cardoso L, Rodrigues D, Gomes L, Carrilho F. Short- and long-term mortality after bariatric surgery: a systematic review and meta-analysis. Diab Obes Metab. 2017;19:1223–32.
Chen Y, Feng Y, Zhang X, Gifford KA, Elmanzalawi Y, Samuels J, et al. Bariatric surgery is associated with reduced incidence of mild cognitive impairment and Alzheimer disease and related dementias: a retrospective cohort study. Ann Surg Open. 2025;6:e541.
Hathaway I, Momodu OE, Williams DM, Beamish AJ, Barry J, Stephens JW. Changes in cognitive function following bariatric surgery: an updated systematic review. Obes Surg. 2024;34:2216–26.
Kim J, Kelley J, Kleinschmit K, Richards N, Adams T. Development of dementia in patients who underwent bariatric surgery. Surg Endosc. 2023;37:3507–21.
Martin-Jiménez CA, Gaitán-Vaca DM, Echeverria V, González J, Barreto GE. Relationship between obesity, Alzheimer’s disease, and Parkinson’s disease: an astrocentric view. Mol Neurobiol. 2017;54:7096–115.
Park KY, Nam GE, Han K, Park HK, Hwang HS. Waist circumference and risk of Parkinson’s disease. NPJ Parkinsons Dis. 2022;8:89.
Benchimol EI, Smeeth L, Guttmann A, Harron K, Moher D, Petersen I, et al. The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement. PLoS Med. 2015;12:e1001885.
Nunes RAB, Neves P, da Costa LMA, Bachour P, Cantarelli MJC, Oliveira GBF, et al. Five-year cardiovascular outcomes in patients with chronic myeloid leukemia treated with imatinib, dasatinib, or nilotinib: a cohort study using data from a large multinational collaborative network. Front Cardiovasc Med. 2023;10:888366.
Benedetto U, Head SJ, Angelini GD, Blackstone EH. Statistical primer: propensity score matching and its alternatives. Eur J Cardio Thorac Surg. 2018;53:1112–7.
Kamal FA, Fernet LY, Rodriguez M, Kamal F, Da Silva NK, Kamal OA, et al. Nutritional deficiencies before and after bariatric surgery in low- and high-income countries: prevention and treatment. Cureus. 2024;16:e55062.
Majumder S, Soriano J, Louie Cruz A, Dasanu CA. Vitamin B12 deficiency in patients undergoing bariatric surgery: preventive strategies and key recommendations. Surg Obes Relat Dis. 2013;9:1013–9.
Al Mansoori A, Shakoor H, Ali HI, Feehan J, Al Dhaheri AS, Cheikh Ismail L, et al. The effects of bariatric surgery on vitamin B status and mental health. Nutrients. 2021;13:1383.
Karlsson HK, Tuominen L, Helin S, Salminen P, Nuutila P, Nummenmaa L. Mesolimbic opioid-dopamine interaction is disrupted in obesity but recovered by weight loss following bariatric surgery. Transl Psychiatry. 2021;11:259.
van der Zwaal EM, de Weijer BA, van de Giessen EM, Janssen I, Berends FJ, van de Laar A, et al. Striatal dopamine D2/3 receptor availability increases after long-term bariatric surgery-induced weight loss. Eur Neuropsychopharmacol. 2016;26:1190–200.
Alosco ML, Spitznagel MB, Strain G, Devlin M, Cohen R, Crosby RD, et al. Improved serum leptin and ghrelin following bariatric surgery predict better postoperative cognitive function. J Clin Neurol. 2015;11:48–56.
Biobaku F, Ghanim H, Monte SV, Caruana JA, Dandona P. Bariatric surgery: remission of inflammation, cardiometabolic benefits, and common adverse effects. J Endocr Soc. 2020;4:bvaa049.
Bray JJH, Foster-Davies H, Salem A, Hoole AL, Obaid DR, Halcox JPJ, et al. Glucagon-like peptide-1 receptor agonists improve biomarkers of inflammation and oxidative stress: a systematic review and meta-analysis of randomised controlled trials. Diab, Obes Metab. 2021;23:1806–22.
Seufert J, Gallwitz B. The extra-pancreatic effects of GLP-1 receptor agonists: a focus on the cardiovascular, gastrointestinal and central nervous systems. Diab Obes Metab. 2014;16:673–88.
Hindsø M, Svane MS, Hedbäck N, Holst JJ, Madsbad S, Bojsen-Møller KN. The role of GLP-1 in postprandial glucose metabolism after bariatric surgery: a narrative review of human GLP-1 receptor antagonist studies. Surg Obes Relat Dis. 2021;17:1383–91.
Cummings JL, Atri A, Feldman HH, Hansson O, Sano M, Knop FK, et al. evoke and evoke+: design of two large-scale, double-blind, placebo-controlled, phase 3 studies evaluating efficacy, safety, and tolerability of semaglutide in early-stage symptomatic Alzheimer’s disease. Alzheimers Res Ther. 2025;17:14.
ClinicalTrials.gov. A Research Study Investigating Semaglutide in People With Early Alzheimer’s Disease (EVOKE Plus) (EVOKE Plus). 2026. https://clinicaltrials.gov/study/NCT04777409.
ClinicalTrials.gov. A Research Study Investigating Semaglutide in People With Early Alzheimer’s Disease (EVOKE). 2026. https://clinicaltrials.gov/study/NCT04777396.
Meissner WG, Remy P, Giordana C, Maltête D, Derkinderen P, Houéto JL, et al. Trial of lixisenatide in early Parkinson’s Disease. N Engl J Med. 2024;390:1176–85.
Vijiaratnam N, Girges C, Auld G, Chau M, Maclagan K, King A, et al. Exenatide once weekly over 2 years as a potential disease-modifying treatment for Parkinson’s disease: protocol for a multicentre, randomised, double blind, parallel group, placebo controlled, phase 3 trial: the ‘Exenatide-PD3’ study. BMJ Open. 2021;11:e047993.
Edison P, Femminella GD, Ritchie C, Nowell J, Holmes C, Walker Z, et al. Liraglutide in mild to moderate Alzheimer’s disease: a phase 2b clinical trial. Nat Med. 2026;32:353–61.
Cummings JL, Atri A, Sano M, Zetterberg H, Scheltens P, Knop FK, et al. Efficacy and safety of oral semaglutide 14 mg (flexible dose) in early-stage symptomatic Alzheimer’s disease (evoke and evoke+): two phase 3, randomised, placebo-controlled trials. Lancet. 2026;407:2167–79.
Tang H, Shao H, Shaaban CE, Yang K, Brown J, Anton S, et al. Newer glucose-lowering drugs and risk of dementia: a systematic review and meta-analysis of observational studies. J Am Geriatrics Soc. 2023;71:2096–106.
Li Q, Jia M, Yan Z, Li Q, Sun F, He C, et al. Activation of glucagon-like peptide-1 receptor ameliorates cognitive decline in type 2 diabetes mellitus through a metabolism-independent pathway. J Am Heart Assoc. 2021;10:e020734.
Cukierman-Yaffe T, Gerstein HC, Colhoun HM, Diaz R, García-Pérez L-E, Lakshmanan M, et al. Effect of dulaglutide on cognitive impairment in type 2 diabetes: an exploratory analysis of the REWIND trial. Lancet Neurol. 2020;19:582–90.
Forst T, De Block C, Del Prato S, Frias J, Lautenbach A, Ludvik B, et al. Novel pharmacotherapies for weight loss: Understanding the role of incretins to enable weight loss and improved health outcomes. Diab Obes Metab. 2025;27:48–65.
van Veldhuisen SL, Gorter TM, van Woerden G, de Boer RA, Rienstra M, Hazebroek EJ, et al. Bariatric surgery and cardiovascular disease: a systematic review and meta-analysis. Eur Heart J. 2022;43:1955–69.
Aminian A, Zajichek A, Arterburn DE, Wolski KE, Brethauer SA, Schauer PR, et al. Association of metabolic surgery with major adverse cardiovascular outcomes in patients with type 2 diabetes and obesity. JAMA. 2019;322:1271–82.
Moussa O, Ardissino M, Heaton T, Tang A, Khan O, Ziprin P, et al. Effect of bariatric surgery on long-term cardiovascular outcomes: a nationwide nested cohort study. Eur Heart J. 2020;41:2660–7.
Courcoulas AP, Daigle CR, Arterburn DE. Long term outcomes of metabolic/bariatric surgery in adults. BMJ. 2023;383:e071027.
Krishnan A, Hadi Y, Alqahtani SA, Woreta TA, Fang W, Abunnaja S, et al. Cardiovascular outcomes and mortality after bariatric surgery in patients with nonalcoholic fatty liver disease and obesity. JAMA Netw Open. 2023;6:e237188–e.
Sjöström L, Peltonen M, Jacobson P, Sjöström CD, Karason K, Wedel H, et al. Bariatric surgery and long-term cardiovascular events. JAMA. 2012;307:56–65.
Wilson RB, Lathigara D, Kaushal D. Systematic review and meta-analysis of the impact of bariatric surgery on future cancer risk. Int J Mol Sci. 2023;24:2023.
Zhang K, Luo Y, Dai H, Deng Z. Effects of bariatric surgery on cancer risk: evidence from meta-analysis. Obes Surg. 2020;30:1265–72.
Wiggins T, Antonowicz SS, Markar SR. Cancer risk following bariatric surgery—systematic review and meta-analysis of national population-based cohort studies. Obes Surg. 2019;29:1031–9.
Aminian A, Wilson R, Al-Kurd A, Tu C, Milinovich A, Kroh M, et al. Association of bariatric surgery with cancer risk and mortality in adults with obesity. JAMA. 2022;327:2423–33.
Syn NL, Cummings DE, Wang LZ, Lin DJ, Zhao JJ, Loh M, et al. Association of metabolic–bariatric surgery with long-term survival in adults with and without diabetes: a one-stage meta-analysis of matched cohort and prospective controlled studies with 174 772 participants. Lancet. 2021;397:1830–41.
Adams TD, Mehta TS, Davidson LE, Hunt SC. All-cause and cause-specific mortality associated with bariatric surgery: a review. Curr Atheroscler Rep. 2015;17:74.
Dickman SL, Himmelstein DU, Woolhandler S. Inequality and the health-care system in the USA. Lancet. 2017;389:1431–41.
Acknowledgements
The authors acknowledge support from the National Institute for Health Research (NIHR) Applied Research Collaboration East Midlands (ARC EM) and the NIHR Leicester Biomedical Research Centre (BRC) and the British Heart Foundation Centre of Research Excellence. SIS acknowledges support from the University of Leicester’s College of Life Sciences PhD Studentship Scheme.
Funding
This study was conducted as part of SIS’s PhD which was funded by the University of Leicester’s College of Life Sciences PhD Studentship Scheme. JG is supported by the Wellcome Trust Leicestershire Healthcare Inequalities Improvement Doctoral Training Programme (223512/Z/21/Z).
Author information
Authors and Affiliations
Diabetes Research Centre, Leicester General Hospital, College of Life Sciences, University of Leicester, Leicester, UK
Suhail I. Shiekh, Dimitris Papamargaritis, Jonathan Goldney, Melanie J. Davies, Kamlesh Khunti, Thomas Yates & Francesco Zaccardi
Leicester Real World Evidence Unit, Diabetes Research Centre, College of Life Sciences, University of Leicester, Leicester, UK
Suhail I. Shiekh, Kamlesh Khunti & Francesco Zaccardi
Department of Diabetes and Endocrinology, Kettering General Hospital NHS Foundation Trust, Kettering, UK
Dimitris Papamargaritis
NIHR Leicester Biomedical Research Centre, Diabetes Research Centre, College of Life Sciences, University of Leicester, Leicester, UK
Jonathan Goldney, Melanie J. Davies & Thomas Yates
TriNetX Europe NV, Sint-Martens-Latem, Belgium
Gema Hernandes
Authors
- Suhail I. Shiekh
- Dimitris Papamargaritis
- Jonathan Goldney
- Gema Hernandes
- Melanie J. Davies
- Kamlesh Khunti
- Thomas Yates
- Francesco Zaccardi
Contributions
SIS, TY, and FZ participated in the design of the study. DP provided clinical advice. FZ supervised the statistical analyses. GH provided advice on the use of the TriNetX platform. SIS performed data analysis, interpreted the data, and wrote the first draft of the manuscript. JG contributed to updating the codelists used for the TriNetX database. SIS, DP, JG, GH, MJD, KK, TY, and FZ contributed to the preparation and revision of the manuscript. SIS, DP, JG, GH, MJD, KK, TY, and FZ read and approved the final manuscript.
Corresponding author
Correspondence to Suhail I. Shiekh.
Ethics declarations
Competing interests
FZ received an honorarium/advisory fee from Menarini, Servier, Daiichi Sankyo, Adelphi Real World. TY is supported by the NIHR Leicester BRC and has received funding from Astra Zeneca for an investigator-initiated project. KK has acted as a consultant, speaker or received grants for investigator-initiated studies for Abbott, Astra Zeneca, Bayer, Novo Nordisk, Sanofi-Aventis, Servier, Lilly and Merck Sharp & Dohme, Boehringer Ingelheim, Oramed Pharmaceuticals, Pfizer, Roche, Daiichi-Sankyo, Applied Therapeutics, Embecta and Nestle Health Science. MJD has acted as a paid consultant, advisory board member, and speaker for Amgen, Boehringer Ingelheim, Eli Lilly, Novo Nordisk, and Sanofi, a paid advisory board member and speaker for AstraZeneca, a paid advisory board member for Abbvie, Roche and Zealand Pharma, and has received grants as an investigator in support of investigator-initiated trials from AstraZeneca, Boehringer Ingelheim, Eli Lilly, Janssen, and Novo Nordisk. GH is a former employee of TriNetX and was working for the company at the time the analysis was conducted. DP has acted as a speaker for Novo Nordisk, Eli Lilly, Boehringer Ingelheim and Johnson and Johnson, as a paid consultant for Regeneron, and has received grants from Novo Nordisk, Novo Nordisk UK Research Foundation, Academy of Medical Sciences/Diabetes UK, Health Education East Midlands and the National Institute for Health and Care Research (NIHR). SIS and JG declare no potential competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Shiekh, S.I., Papamargaritis, D., Goldney, J. et al. Bariatric surgery and risk of dementia and Parkinson’s disease in patients with obesity. Int J Obes (2026). https://doi.org/10.1038/s41366-026-02222-z
Received: 20 July 2025
Revised: 18 August 2026
Accepted: 07 September 2026
Published: 07 October 2026
Version of record: 07 October 2026
DOI: https://doi.org/10.1038/s41366-026-02222-z
来源:Nature 健康护理主题 · nature.com