Previous meta-analyses have shown that metabolic surgery reduces mortality as well as major adverse cardiovascular events. However, it remains underutilized as a cardiovascular intervention. This study aims to identify patient subgroups most likely to benefit as well as compare outcomes from different metabolic surgery procedures . We systematically searched online databases for comparative cohorts (surgery vs no surgery) with ≥4-year follow-up and >200 participants. The primary outcome was all-cause mortality; secondary outcomes were cardiovascular mortality, myocardial infarction, heart failure (HF), stroke, and new atrial fibrillation (AF). Only unadjusted hazard ratios (HR) were extracted using random-effects models, and meta-regression was performed to identify any significant predictors. We included 25 studies (n = 659,517). MS was associated with lower risk of mortality (18 studies; HR 0.55 (95% CI [0.45, 0.67], p < 0.001), cardiovascular mortality (4; 0.36, [0.18, 0.72], p = 0.004), myocardial infarction (6; 0.61 [0.41, 0.90], p = 0.01), new HF (7; 0.55 [0.47, 0.65], p < 0.001), stroke (6, 0.80 [0.66 to 0.96], p = 0.02), and AF (3, 0.70 [0.55 to 0.89], p = 0.003). There were no specific study-level predictors that modified all-cause mortality. Associations were seen for cardiovascular mortality (ischemic heart disease, HF, and smoking) and AF (younger age, smoking). Benefits across all end points were consistent regardless of procedure subtypes. In conclusion, MS confers broad cardiovascular benefit with minimal effect modification from specific patient factors. Individual patient-level analyses and comparisons with modern incretin therapies are warranted.
Clinical Perspective
What is new? This meta-analysis of 25 studies (n = 659,517) using exclusively unadjusted hazard ratios demonstrates consistent reductions in all-cause mortality and major cardiovascular outcomes following metabolic surgery, with no significant differences between surgical procedures and limited effect modification by patient characteristics. What are the clinical implications? These findings support broader referral for metabolic surgery irrespective of procedure type, reinforce its role as a cardiovascular risk-reducing intervention, and provide clinicians with clearer evidence to counsel patients regarding expected cardiovascular benefits.
Research Perspective
What new question does this study raise? The observation that most patient-level factors do not substantially modify mortality benefit raises the question of whether individual-level clinical, metabolic, or imaging markers better predict cardiovascular response to metabolic surgery. What question should be addressed next? Future studies should directly compare metabolic surgery with contemporary incretin-based pharmacotherapy and perform individual patient meta-analyses to identify precise predictors of benefit and optimize personalized obesity-related cardiovascular risk reduction strategies.
Obesity is a global health crisis, with rapidly rising prevalence and substantial cardiovascular burden. It is strongly associated with coronary heart disease, heart failure (HF), hypertension, stroke, atrial fibrillation (AF), and sudden cardiac death, with cardiovascular mortality being the primary driver of obesity-related deaths. Metabolic surgery, including Roux-en-Y gastric bypass, sleeve gastrectomy, biliopancreatic diversion, and gastric banding––is currently the most effective strategy for sustained weight loss and provides additional metabolic benefits independent of weight loss. Metabolic surgery has been shown to reduce the incidence of major adverse cardiovascular events (MACE). Despite these benefits, utilization remains low, with only 1% of eligible individuals in the United States undergoing metabolic surgery. Identifying which patients derive the greatest cardiovascular benefit is essential to guide referral and optimize resource allocation. Previous studies suggest enhanced survival benefit in certain populations, including younger individuals, males, and those with diabetes. , However, existing meta-analyses often combine adjusted estimates with heterogeneous covariate models and rarely explore procedure-specific outcomes or study-level effect modifiers. ,, This systematic review and meta-analysis therefore aimed to evaluate the association between metabolic surgery and major cardiovascular outcomes using exclusively unadjusted hazard ratios (HR), as well as examining whether study-specific characteristics and surgical subtypes influenced treatment effects.
Methods
A systematic literature review of Medline, Embase, and Web of Science was performed. The search was performed until July 24, 2025. The databank was supplemented with research papers identified from the reference list of relevant studies. All studies were then reviewed for inclusion by two investigators (A.S, M.H) with any discrepancy mediated by a third investigator (N.N). Duplicate and non-English-language studies were excluded, and exclusion subsequently occurred at the title, abstract, or full-text level.
Study selection
We included study designs that were intervention-comparator trials. This incorporated randomized controlled trials, prospective or retrospective cohort studies. Studies were included if they compared individuals with obesity undergoing metabolic surgery as the intervention. The comparator cohort was individuals with obesity who did not receive surgery, and either had lifestyle modification or no formalized therapy, that is, not receiving medical weight loss therapy. Study inclusion also required reporting of either the primary outcome (all-cause mortality) or at least one of the secondary outcomes. These included cardiovascular mortality, myocardial infarction (MI), (both fatal and nonfatal), new HF, HF hospitalization, stroke (ischemic, but excluding transient-ischemic attack), or new AF. We restricted studies to >4-year follow-up and >200 total participants. Studies were also only included if they were published within the last 20 years (2005 onward) in an effort to reflect the most up-to-date safety profile and surgical techniques. We elected to exclude cohorts that were restricted to only chronic kidney disease/end-stage renal failure, or nonalcoholic fatty liver disease groups. If studies had overlapping datasets for the same outcome, the data of the study with the longest follow-up were included. If studies contained overlapping datasets for different outcomes, these studies were only analyzed for specific outcomes extracted once (informed by the longest follow-up period).
Data extraction
The data extracted from each study included baseline demographics and the number of events and unadjusted HRs for each of the outcomes mentioned above. In the situation where only adjusted data was available, an unadjusted HR was calculated based on the raw data if the cumulative incidence and numbers at risk at different time intervals were demonstrated in either the published article or its accompanying supplementary material using the HR calculator produced by Tierney et al. , If the unadjusted HR was either not reported or could not be calculated in this manner, the study was excluded. Standard error was calculated from reported confidence intervals. While the safety of metabolic surgery was not the focus of this meta-analysis, safety outcomes were opportunistically extracted from included studies where possible.
End points
The primary end point was all-cause mortality. Secondary end points included cardiovascular mortality, MI, H, stroke, and new AF. Our aim was to quantify the association between metabolic surgery and each end point using unadjusted HRs and to assess effect modification by study-level covariates and surgery type.
Quality assessment
The Newcastle-Ottawa Quality Assessment Scale for Cohort Studies was used for study quality to examine the risk of bias of each study ( Supplement Table 1 ).
Statistical analysis
RStudio (RStudio, Inc, Boston, MA; version 2023.03.0) was used for statistical analysis. Continuous variables with a Gaussian distribution are presented as mean ± standard deviation (SD), with ordinal variables presented as counts (percentages). HRs were log-transformed and pooled using DerSimonian—Laird random-effects modeling. Heterogeneity was assessed using I 2, Q, τ², and H² statistics. Meta-regression explored study-level moderators through mixed-effects meta-regression models, and model parameters included beta coefficients, standard errors, and p-values for each moderator. Where possible, results were visualized using bubble plots. Sensitivity analysis was conducted through leave-one-out analysis for each outcome, and subgroup analysis was performed stratifying studies according to metabolic surgery type. Publication bias was assessed using funnel plots, Egger’s regression test, trim-and-fill analyses, and Galbraith plots.
Results
Search results
Our search strategy yielded 3,808 studies, and after screening, we included 25 studies in the current review, including 659,517 patients ( Figure 1 ). Only prospective and retrospective cohort studies were included as there are currently no published randomized controlled trials examining the association of metabolic surgery and cardiovascular outcomes. Study demographics are outlined in Table 1 . Of note, Courcoulas et al. and Lent et al. both had 2 independent cohorts, and have therefore each been presented as separate cohorts during statistical analysis. Across included studies, baseline characteristics—where reported ( Table 1 )—included a mean age of 46.2 years (range 36.3 to 54.5) and a mean BMI of 43.2 kg/m 2 (range 36.6 to 47.4). Cohorts comprised 27.8% men; comorbidities included diabetes (30.0%), hypertension (42.9%), hyperlipidemia (42.4%), CKD (8.2%), AF (3.5%), ischemic heart disease (IHD) (4.0%), HF (4.6%), obstructive sleep apnea (12.3%), chronic lung disease (12.5%), depression (22.3%), and smoking (29.9%). Of those in the metabolic surgery arm, 66.7% underwent RYGB, 20.0% SG, 8.2% GB, and 5.7% other. Mean follow-up was 4.8 years in controls and 8.9 years in the intervention group.
PRISMA diagram for study selection.
Table 1
Characteristics of included studies, including variables extracted and outcomes analyzed
| Study | Control group | Intervention | Study design | Cohort | Major inclusion criteria | Major exclusion criteria | Data extracted | Follow-up (years) | |||
|---|---|---|---|---|---|---|---|---|---|---|---|
| N | Variables extracted | Surgery type | N | Variables extracted | |||||||
| Adams et al. | 21,837 | Age (42.3 ± 11.9), BMI (46.2), Male (20.9%), | 69.2% RYGB, 14% SG, 12% GB, 4.8% other | 21,837 | Age (42.2 ± 11.7), BMI (45 ± 8.3), Male (20.9%), | R | Utah Population Database, 1982 to 2018, Matched | Severe obesity or underwent metabolic surgery | Age <18 or >80 BMI <30 Could not be matched | All-cause mortality, cardiovascular mortality | Control: 13.2 Surgical: 13.3 |
| Aminian et al. | 11,435 | Age (54.5 ± 12), BMI (43 ± 5.8), T2DM (100%), Male (35.8%), AF (6.1%), IHD (9.6%), HF (11.7%), smoking (52.5%), HTN (74.9%) | 63% RYGB, 32% SG, 5% GB, 0.2% other | 2,287 | Age (52.5 ± 12.1), BMI (45.6 ± 8.7), T2DM (100%), AF (6.6%), IHD (10.4%), HF (10.4%), smoking (43.6%), HTN (85.4%) | R | Cleveland Clinic Health System, USA, 1998 to 2018, Matched | T2DM before BS, Age 18 to 80, BMI ≥30, HbA1c ≥6.5%, or ≥1 diabetic drug | Missing info, <18 or >80 years, prior cancer, EF <20%, solid organ transplant | All-cause mortality, new HF, stroke, AF | Control: 4 Surgical: 3.3 |
| Arterburn et al. | 7,462 | Age (53 ± 8.7), BMI (46 ± 7.3), Male (74.0%), Diabetes (55.0%), IHD (18.0%), HTN (70.0%), | 74% RYGB, 15% SG, 10% GB, 1% other | 2,500 | Age (52 ± 8.8), BMI (47 ± 7.9), Diabetes (55.0%), IHD (20.0%), HTN (80.0%), | R | Veteran Affairs Surgical Quality Improvement Program data, 2000 to 2011, Matched | Underwent metabolic surgery | Missing info, BMI <35, recent cancer, IBD, ESRD, pregnant, ascites | All-cause mortality | Control: 6.6 Surgical: 6.6 |
| Benotti et al. | 1,724 | Age (45.1 ± 10.6), BMI (46.5 ± 6.1), Male (13.0%), Diabetes (28.0%), AF (0%), IHD (0%), HF (0%), smoking (42.0%), | 100% RYGB | 1,724 | Age (45 ± 10.6), BMI (46.5 ± 6), Male (13.0%), Diabetes (28.0%), AF (0%), IHD (0%), HF (0%), smoking (42.0%), | R | Geisinger Health Center, USA, 2002 to 2012, Matched | Age 20 to 80 years, BMI >35, no pre-existing CVD (IHD, CM, arrhythmias, HF, stroke, PAD) | Missing CV risk factor data | MI, HF, stroke | Control: 5.7 Surgical: 5.8 |
| Carlsson et al. | 2,040 | Age (48.7 ± 6.3), BMI (40.1 ± 4.7), T2DM (12.9%), Male (29.1%), smoking (20.7%), HTN (63.8%) | 69.2% RYGB, 14% SG, 12% GB, 4.8% other | 2,007 | Age (47.2 ± 5.9), BMI (42.4 ± 4.5), Male (29.2%), T2DM (12.9%), smoking (25.8%), HTN (78.3%) | P | Swedish Obese Subjects Registry, 1987 to 2001, Matched | Age 37 to 60, BMI >34 men, >38 women, | Earlier gastric/duodenal surgery, ongoing malignancy, MI <6 months ago, drug/alcohol | All-cause mortality | Control: 24 Surgical: 22 |
| Ceriani et al. | 1,405 | Age (43.5 ± 12.5), BMI (46.8 ± 3.8), Male (30.0%), Diabetes (27.4%), IHD (3.6%), | 100% other | 472 | Age (43.1 ± 10.61), BMI (47.3 ± 7.46), Male (25%), Diabetes (23.5%), IHD (3.2%) | R | LAGB10 study group, Italy, 1999 to 2008, Matched | Referred to obesity clinic, aged 18 to 65 years, and with BMI ≥40 or ≥35 with comorbidities | – | All-cause mortality | 12.1 |
| Courcoulas et al. (RYGB) | 65,416 | Age (43.6 ± 11.6), BMI (44.2 ± 6.6), Male (18.5%), Diabetes (39.6%), IHD (2.8%), smoking (33.0%), HTN (50.0%) | 100% RYGB | 17,258 | Age (45.7 ± 11), BMI (45.1 ± 7.1), Male (17.9%), Diabetes (39.7%), IHD (2.5%), smoking (34.1%), HTN (60.0%) | R | Kaiser Permanente regions in Washington and California, 2005 to 2015 Matched | Age 19 to 79 years, BMI ≥35, primary metabolic surgery | <1 year of enrolment, pregnancy, cancer (except non-melanoma skin cancer) | All-cause mortality | – |
| Courcoulas et al. (SG) | 39,795 | Age (44.9 ± 11.7), BMI (43 ± 6.2), Male (19.6%), Diabetes (25.9%), IHD (1.9%), smoking (32.5%), HTN (41.9%) | 100% SG | 13,900 | Age (44.6 ± 11.1), BMI (43.6 ± 6.5), Male (19.1%), Diabetes (26.1%), IHD (1.8%), smoking (35.1%), HTN (49.3%) | R | |||||
| Doumouras et al. | 3,455 | Age (52.4 ± 9.7), BMI (44.1 ± 8.3), T2DM (100%), Male (28.4%), AF (1.6%), IHD (4.4%), HF (3.3%), smoking (11.8%), HTN (11.6%) | 86.7% RYGB, 13.3% SG | 3,455 | Age (51.7 ± 9.2), BMI (45.3 ± 7.6), T2DM (100%), Male (28.4%), AF (2.2%), IHD (6.2%), HF (2.9%), smoking (8.1%), HTN (15.8%) | R | Ontario Bariatric Network, 2010 to 2016, Matched | Not specified other than underwent metabolic surgery | Non-Ontario patients, age >70 years, BMI <35, cancer, substance abuse, palliative care, pregnancy, organ transplantation, liver/heart disease, revascularization <6 months ago, ascites <1 year ago | All-cause Mortality, cardiovascular mortality | Control: 4.6 Surgical: 4.6 |
| Eliasson et al. | 6,132 | Age (50.5 ± 12.7), BMI (42.4 ± 5.7), Male (38.6%), Diabetes (100%), IHD (4.3%), HF (4.1%), smoking (17.1%), | 100% RYGB | 6,132 | Age (48.5 ± 9.8), BMI (42.0 ± 5.7), Male (38.6%), Diabetes (100%), IHD (3.8%), HF (2.8%), smoking (8.8%) | P | National Diabetes Registry and Scandinavian Obesity Surgery Registry, 2007 to 2014, matched | Undergone RYGB in Swedish Hospitals with complete socioeconomic data | – | Mortality, CV mortality, MI, | 3.5 |
| Jamaly et al. | 2,021 | Age (48.6 ± 6.2), BMI (40.1 ± 4.7), Diabetes (12.7%), Male (28.8%), smoking (20.9%), HTN (63.6%), | – | 2,000 | Age (47.2 ± 5.9), BMI (42.4 ± 4.5), Diabetes (17.2%), Male (29.3%), smoking (21.1%), HTN (78.3%), | P | Swedish Obese Subjects Registry, 1987 to 2001, Matched | Age 37 to 60 years, male BMI >34, female BMI >38 | MI <6 months, earlier gastric surgery, drugs/alcohol, Diagnosis of AF, | AF | 22 |
| Jamaly et al. | 2,030 | Age (48.6 ± 6.2), BMI (40.1 ± 4.7), Diabetes (12.7%), Male (28.8%), smoking (20.9%), HTN (63.6%), | – | 2,003 | Age (47.2 ± 5.9), BMI (42.4 ± 4.5), Diabetes (17.2%), Male (29.2%), smoking (25.8%), HTN (78.4%), | P | MI <6 months, earlier gastric surgery, drugs/alcohol, Diagnosis of HF, | HF | 19 | ||
| Lent et al (Non-diabetic) | 1,803 | Age (43.9 ± 11), BMI (47.3 ± 6.4), T2DM (0%), Male (13.0%), smoking (49.0%), | 100% RYGB | 1,803 | Age (43.8 ± 11), BMI (47.4 ± 6.4), T2DM (0%), Male (13.0%), smoking (63.0%) | R | Geisinger Health Center, USA, 2004 to 2015, Matched | Age 18 to 70 years, BMI >40 (or >35 + 1 comorbidity), active in primary care, no serious mental health disorder, no illicit substance abuse | SG, GB, other metabolic surgery, diabetic | Mortality | Control: 6.7 Surgery:- |
| Lent et al (Diabetic) | 625 | Age (52.5 ± 9.4), BMI (44.9 ± 6.1), T2DM (100%), Male (27.0%), smoking (59.0%), | 100% RYGB | 625 | Age (52.5 ± 9.4), BMI (44.9 ± 6.0), T2DM (100%), Male (27.0%), smoking (71.0%) | R | SG, GB, other metabolic surgery, non-diabetic | Control: 5.8 Surgery:- | |||
| Liakopoulos et al. | 5,321 | Age (47.1 ± 11.5), BMI (40.9 ± 7.3), T2DM (100%), Male (36.2%), IHD (5.9%), HF (3.1%), smoking (17.7%), | 100% RYGB | 5,321 | Age (49.0 ± 9.5), BMI (42.0 ± 5.7), T2DM (100%), Male (39.4%), IHD (7.4%), HF (2.7%), smoking (10.8%), | P | National Diabetes Register and Scandinavian Obesity Surgery Registry, 2007 to 2015, Matched | Age 18 to 65 years, T2DM, primary RYGB | Metabolic surgery other than RYGB, non-T2DM | HF hospitalisations | Control: 4.6 Surgical: 4.7 |
| Migliore et al. | 1,954 | Age (46.4 ± 11), BMI (46.9 ± 7.4), Male (33.7%), Male (28.0%), smoking (28.0%), | 18.7% RYGB, 64.4% SG, 16.9% GB | 331 | Age (40.7 ± 11.3), BMI (47.4 ± 6.6), Male (30.0%), Diabetes (18.7%), smoking (25.4%), | R | Istituto Auxologico Italiano (IAI) Piancavallo, 2002 to 2018 Non-matched | Admitted for severe obesity or its complications, age 18 to 60 years, BMI≥40 | – | All-cause mortality | 10.2 |
| Moussa et al. | 3,701 | Age (36.3 ± 11.1), BMI (40.3), Male (20.2%), Diabetes (23.9%), smoking (36.6%), HTN (49.2%) | 38% RYGB, 14% SG, 35% GB, 0.6% Other | 3,701 | Age (36.3 ± 11.1), BMI (40.5), Male (20.2%), Diabetes (25.0%), smoking (37.0%), HTN (52.1%) | R | UK Clinical Practice Research Datalink, 1987 to 2020, Matched | Not specified other than metabolic surgery | BMI <35, MACE before index date, lost to follow-up <12 months after index date, missing data: age, BMI, sex | Stroke | 11.7 |
| Pirlet et al. | 116 | Age (52.1 ± 8.4), BMI (41.2 ± 6.7), Male (74.1%), Diabetes (50.9%), AF (4.3%), IHD (100%), HF (7.8%), smoking (75.9%), HTN (81.0%) | 2.6% RYGB, 58% SG, 39.4% other | 116 | Age (52.9 ± 7.8), BMI (42.0 ± 6.1), Male (70.7%), AF (2.6%), IHD (100%), HF (5.2%), smoking (72.4%), HTN (81.0%) | P | Quebec Heart and Lung Institute 1992 to 2017 Matched | Underwent metabolic surgery and had history of PCI/CABG | For controls: Coronary event within (unspecified) time period | All-cause, MI, stroke | 9.8 |
| Reges et al. | 25,155 | Age (40.3 ± 4.8), BMI (40.3 ± 4.8), Male (34.5%), Diabetes (28.5%), IHD (10.6%), smoking (29.5%), HTN (42.4%), | 16.6% RYGB, 40% SG, 43.3% GB | 8,385 | Age (45.7 ± 12.6), BMI (40.6 ± 3.9), Male (34.5%), IHD (11.8%), smoking (38.0%), HTN (44.4%) | R | Clalit Health Service, Israel, 2005 to 2014, Matched | Age <24 years, membership Clalit health service | Missing BMI, BMI <30, pregnancy, active cancer, IBD, ESRD, ascites) | All-cause mortality | Control: 4.0, Surg: 4.3 |
| Satarehdan et al. | 6,725 | Age (40.3 ± 14.1), BMI (40.5 ± 8.1), Male (23.2%), T2DM (18.6%), HTN (23.4), | 28.7% RYGB, 21.5% SG, 49.8% Other | 6,190 | Age (38 ± 11.1), BMI (44.3 ± 6.7), Male (20.5%), T2DM (21.7%), HTN (20.9%), | R | Iran National Obesity Surgery Database, 2009 to 2019, Matched | Aged 20 to 75 years | Missing age | All-cause mortality | Control: 4.8, Surgery: 4 |
| Singh et al. | 9,995 | Age (45.3 ± 10.5), Male (18.9%), T2DM (20.9%), AF (1.4%), IHD (3.3%), HF (0.8%), smoking (41.9%), HTN (29.5%) | 38.9% RYGB, 22.4% SG, 38% GB, 0.7% other | 5,170 | Age (45.2 ± 10.6), T2DM (22.7%), Male (19.6%), AF (1.6%), IHD (3.0%), HF (0.8%), smoking (44.0%), HTN (31.0%), | R | Improvement Network (THIN) 1990 to 2018 | >1 year registered in general practice | BMI <30, age >75 years, gastric cancer, gastric balloon, endo-barrier, or revisional bariatric surgery | AF | 3.9 |
| Sjostrom et al. | 2,037 | Age (47.4 ± 6.1), BMI (40.9 ± 4.3), Male (28.9%), Diabetes (6.1%), smoking (20.2%) | – | 2,010 | Age (46.1 ± 5.8), BMI (41.8 ± 4.4), Male (29.4%), Diabetes (7.4%), smoking (29.4%) | P | Swedish Obese Subjects Registry, 1987 to 2001, Matched | Age 37 to 60 years | Prior gastric/duodenal surgery, ongoing malignancy, MI <6 months ago, drugs/alcohol | CV mortality | 14.7 |
| Sundstrom et al. | 13,701 | Age (41.5), BMI (41.4), AF (1.0%), IHD (0.8%), smoking (17.1%) | 100% RYGB | 25,804 | Age (41.3), BMI (41.5), AF (1.0%), IHD (1.1%), smoking (16.2%) | R | Scandinavian Obesity Surgery Registry and Itrim Health Database, 2007 to 2012 | Aged >18 years, BMI 30 to 49.9, underwent RYGB, no baseline heart failure | Crossed over from Itrim to Scandinavian Obesity Surgery Registry, missing education or marital status info | New HF | 4.1 |
| Wiebe et al. | 2,97,945 | Age (48.3 ± 16.1), Male (30.2%), Diabetes (18.9%), AF (3.9%), IHD (3.7%), HF (4.7%), HTN (38.6%) | – | 62,12 | Age (43.7 ± 10.5), Male (14.8%), Diabetes (29.4%), AF (2.2%), IHD (1.0%), HF (2.3%), HTN (47.6%) | R | Alberta Kidney Disease Network database 1997 to 2018 | Underwent metabolic surgery, severe obesity (BMI >35 + >1 comorbidity or >40) | Hospitalised at the same time for GI/abdominal cancer or perforate GI ulcer as undergoing metabolic surgery | All-cause death | – |
| Wong et al. | 1,399 | Age (51.0 ± 13.4), BMI (36.6 ± 6.5), Male (46.2%), T2DM (100%), IHD (0%), HF (0%), HTN (78.3%), | 16.2% RYGB, 80.5% SG | 303 | Age (51.4 ± 12.3), BMI (37.4 ± 5.0), Male (45.9%), T2DM (100%), IHD (0%), HF (0%), HTN (78.5%) | R | Hospital Authority Database, Hong Kong, 2006 to 2017, Matched | T2DM | BMI <27.5, non-T2DM, history of CVD, eGFR<30 | All-cause mortality, MI, HF, stroke | 2.6 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
