AIMS:The comparative association of sodium-glucose cotransporter-2 inhibitors (SGLT2i) and glucagon-like peptide-1 receptor agonists (GLP-1RA) with incident atrial fibrillation (AF) or atrial flutter remains uncertain, particularly with early use of both classes. We compared three initial treatment strategies in adults with type 2 diabetes. METHODS:We emulated a three-arm target trial using population-based healthcare data from 2015-2025. Participants were assigned to SGLT2i without GLP-1RA, GLP-1RA without SGLT2i, or early combined therapy according to dispensings during a 30-day treatment-assignment period. Follow-up began at the end of this period using a landmark design. Treatment groups were balanced using inverse probability of treatment weighting, and cumulative incidence was estimated using weighted Aalen-Johansen methods accounting for competing mortality. RESULTS:Among 216,293 participants, 114,572 received SGLT2i without GLP-1RA, 91,524 received GLP-1RA without SGLT2i, and 10,197 received early combined therapy. At 5 years, cumulative AF or atrial flutter incidence was 4.3%, 4.8%, and 4.2% among participants receiving SGLT2i without GLP-1RA, GLP-1RA without SGLT2i, and early combined therapy, respectively. SGLT2i without GLP-1RA was associated with lower risk than GLP-1RA without SGLT2i (RR 0.88, 95% CI 0.83-0.94), while risk was similar between SGLT2i without GLP-1RA and combined therapy (RR 1.00, 95% CI 0.84-1.19). CONCLUSIONS:Among adults with type 2 diabetes, treatment strategies incorporating SGLT2i were associated with a lower risk of incident AF or atrial flutter than GLP-1RA without SGLT2i. These findings extend the cardiovascular evidence supporting SGLT2i and provide new comparative data regarding early combined use of SGLT2i and GLP-1RA.
OBJECTIVE:Rheumatoid arthritis (RA) confers substantially elevated cardiovascular disease (CVD) risk. We quantified the association between joint risk factor control and incident CVD among RA patients and examined modification by genetic susceptibility. METHODS:In this prospective UK Biobank cohort, 4,426 RA patients and 17,704 matched non-RA controls were followed for CVD, coronary heart disease (CHD), stroke, and heart failure. A nine-component risk factor control score (0-9) was constructed, incorporating healthy diet, smoking, physical activity, blood pressure, glycated hemoglobin, low-density lipoprotein cholesterol, high-sensitivity C-reactive protein, body mass index, and estimated glomerular filtration rate. Cox models, sensitivity analyses, and restricted mean survival time (RMST) were applied. Least absolute shrinkage and selection operator (LASSO) regression identified CVD-predictive metabolomic features; mediation analysis quantified the metabolomic risk score (MRS) contribution. Polygenic risk score (PRS) joint analyses assessed genetic effect modification. RESULTS:Better control was associated with graded CVD risk reduction (moderate: HR 0.77, 95% CI 0.64-0.93; high: HR 0.46, 95% CI 0.33-0.64 vs. low). High control yielded CVD and CHD risks comparable to non-RA controls and lower stroke risk (HR 0.34, 95% CI 0.13-0.91). RMST showed +3.58 CVD-free years (95% CI 2.49-4.68; P < 0.001). Mediation indicated partial MRS mediation (indirect proportion 34.1%; 20.8% adjusted). No significant interaction with MRS; high control attenuated risk even in high genetic risk. CONCLUSION:Comprehensive control attenuates and may offset RA-excess CVD risk, even with high genetic susceptibility. The effect is partially metabolically mediated yet independent of metabolic status, supporting dual-pathway prevention.
AIMS:Current guidelines suggest a unique 10-year cardiovascular disease (CVD) risk threshold for the oldest to consider lipid-lowering treatment for CVD prevention. We aimed to assess the benefit of initiating primary prevention with statins in non-diabetic adults aged 70-89 according to their SCORE2-OP, and to explore risk thresholds for different ages. METHODS:We conducted an observational retrospective cohort study with the Catalan primary care database (SIDIAP) with an intention-to-treat approach. We assessed incident CVD (myocardial infarction and stroke), all-cause mortality, and adverse effects. Overall and age-stratified propensity-adjusted Cox proportional hazard models were fitted, using restricted cubic splines for the treatment interaction with SCORE2-OP. RESULTS:We analyzed 167,360 individuals (mean age 77.70 years; 36.9% men). The mean baseline SCORE2-OP risk was 13.81%. Overall, treatment was not associated with a CVD reduction in lower risks, with a protective association becoming stable around a SCORE2-OP of 16% (HR=0.90 [0.84-0.96]). The different age groups reached similar protective associations (HR ≈ 0.85), but at different estimated risks. A mild significant increase of type 2 diabetes was observed in statin initiators (1.07 [1.02-1.12]). We did not observe a protective association with all-cause mortality across the whole estimated risk spectrum. CONCLUSIONS:Age-specific risk thresholds of the SCORE2-OP estimated risk may be further developed and considered when weighing the benefits of initiating primary prevention of incident CVD with statins in non-diabetic adults aged 70+. Our results do not support treatment initiation of primary prevention to reduce all-cause mortality at any level of estimated CVD risk.
AIM:The SCORE2/SCORE2-OP, recommended for estimating 10-year risk of first atherosclerotic cardiovascular disease (ASCVD), does not include lifestyle factors. We investigated whether adding self-reported physical activity and Mediterranean diet adherence improves prediction beyond SCORE2/SCORE2-OP. METHODS:We included participants from the population-based CoLaus|PsyCoLaus cohort with data for the Physical Activity Frequency Questionnaire (PAFQ) and Mediterranean Diet Score (MedDietScore). We fitted Cox-regression models including SCORE2 or SCORE2-OP as covariates, depending on the participants' age, alone or in addition to PAFQ, MedDietScore or both. We assessed discrimination (Harrell's C), calibration slope, and net reclassification improvement (NRI) for all models. RESULTS:A total of 2,587 participants (59% women) were followed for a median duration of 9.0 years (IQR: 8.13-11.6), during which 145 ASCVD events occurred. 2,348 were younger than 70 years and evaluated using SCORE2, with 106 ASCVD events. The remaining 239 were aged 70 years or older and assessed with SCORE2-OP, with 39 ASCVD events. Adding physical activity or diet or both to SCORE2 did not improve discrimination (Harrell's C: 0.74 for all models), calibration (slope: 1.06 for all), or reclassification (continuous total NRI: -8.8 to 2.5%). Adding both lifestyle factors to SCORE2-OP improved discrimination (Harrell's C: 0.63 to 0.66), calibration (slope: 1.34 for SCORE2-OP to 1.16 for full model) and reclassification (continuous total NRI: 41% (95%CI 7-74%) with event NRI of 28% and non-event NRI of 13%). CONCLUSION:The addition of self-reported physical activity and Mediterranean diet may provide modest incremental value to improve ASCVD risk prediction in older adults beyond the traditional SCORE2-OP.
AIMS:Stroke remains a major global health challenge, with high systolic blood pressure (HSBP) a leading modifiable contributor. The hypertension care-cascade gap most strongly associated with HSBP-attributable stroke burden across countries remains unclear. METHODS:This country-level ecological study used NCD Risk Factor Collaboration hypertension care-cascade estimates, Global Burden of Disease 2023 HSBP-attributable stroke estimates, and Socio-demographic Index (SDI) data, 1990-2019, including 199 countries/territories and 5970 country-years. Indicators were diagnosis, treatment, blood pressure control, and cumulative gaps; outcomes were HSBP-attributable stroke disability-adjusted life-year (DALY) and stroke death rates. Analyses included correlations, country-year models, sex-age-stratified models, exploratory 10-year age-shifted models, and 2019 country typology. RESULTS:By 2019, 80.2% of people with hypertension lacked blood pressure control. The blood pressure control gap aligned most clearly with HSBP-attributable stroke burden; each 10-percentage-point higher gap was associated with a 49.1% higher stroke DALY rate (95% CI 38.5-60.5) and a 47.7% higher stroke death rate (37.9-58.2), exceeding other gaps. Associations were consistent across sex-age strata. Priority analyses identified low-SDI older men as high-gap/high-burden strata, while exploratory age-shifted analyses showed the strongest aggregate associations in men aged 30-44 years and women aged 35-44 years at the index year. CONCLUSION:The control gap showed the strongest country-level alignment with HSBP-attributable stroke burden, with high-gap/high-burden patterns concentrated among older men in low-SDI settings. Exploratory age-shifted patterns in younger adults warrant further study. These findings highlight the potential value of the control gap for population-level surveillance and comparative assessment.
BACKGROUND:Low cardiorespiratory fitness (CRF) and traditional cardiovascular risk factors are associated with mortality, but their relative contributions to mortality and the independent association of CRF remain incompletely understood. METHODS:We analyzed a large Veterans Administration cohort undergoing maximal exercise testing (n=580,290; mean age 65.1±9.7 years) and followed them for 10.2±5.1 years. Relative risks (RRs) for mortality were estimated using Poisson regression with robust (HC3) standard errors across nested models adjusting sequentially for age, sex, BMI, diabetes, smoking, and statin therapy. Population attributable risks (PARs) and exposure impact numbers (EINs) were calculated for low CRF and other major risk factors. RESULTS:Individuals in the highest CRF quintile had approximately 74% lower mortality vs. those in the lowest quintile (<5 METs). Hypertension had the greatest population-level impact (PAR 21.8%; 95% CI 15.2-28.0), followed by low CRF (PAR 14.7%; 95% CI 8.1-21.1). In a fully adjusted sensitivity analysis, low CRF demonstrated the strongest independent association with mortality (RR 1.77; 95% CI 1.75-1.79) compared with hypertension (RR 1.33; 95% CI 1.32-1.35). Absolute EIN was lowest for low CRF (2.36), suggesting that for every 100 individuals whose fitness improved over the observed follow-up period, approximately 24 fewer deaths would be expected. CONCLUSIONS:Low CRF and hypertension were the strongest predictors of mortality and accounted for the greatest proportion of population-level risk. These findings support CRF as a modifiable clinical vital sign and highlight the complementary importance of fitness promotion and blood pressure control for reducing mortality.
AIMS:This study aimed to explore the associations of fine particulate matter (PM2.5) components, residential greenness, and heatwaves with incident venous thromboembolism (VTE). METHODS:This study included 483,610 participants from the UK Biobank (mean age 56.5 years; 54.3% female). Exposures included PM2.5 components (elemental carbon, organic matter, sulfate, ammonium, and nitrate) estimated at a 3×3 km resolution, residential greenness measured by Normalised Difference Vegetation Index within 300 m buffers, and heatwaves defined using multiple temperature thresholds. The outcome was identified through hospital admission and death records. Cox proportional hazards models along with interaction and mediation analyses were fitted. RESULTS:During a median follow-up of 11.83 years, 9,988 incident VTE cases were identified. Each interquartile range increase in exposure to elemental carbon, sulfate, and ammonium was positively associated with VTE risk, with hazard ratios (HRs) and 95% confidence intervals (CIs) of 1.052 (1.005, 1.101), 1.064 (1.001, 1.130), and 1.090 (1.010, 1.176), respectively. Elemental carbon showed the strongest association among these components. Each interquartile range increase in residential greenness was associated with lower VTE risk (HR: 0.944, 95% CI: 0.916, 0.972), with specific PM2.5 components (elemental carbon and ammonium) partly mediating this association. Under different heatwave definitions, the HRs (95% CIs) for incident VTE ranged from 1.149 (1.137, 1.161) to 1.407 (1.396, 1.417). Significant additive interactions with heatwaves under different definitions were identified for sulfate and ammonium, but not consistently for other PM2.5 components or residential greenness, on VTE risk. Genetic factors, assessed by a polygenic risk score, showed significant additive interactions with ammonium and nitrate, as well as a multiplicative interaction with residential greenness, suggesting that genetic susceptibility may modify these associations. CONCLUSIONS:This study identified that exposure to specific PM2.5 components, low residential greenness, and heatwaves were associated with elevated VTE risk.
AIMS:Sudden cardiac arrest (SCA) often occurs in individuals without diagnosed cardiovascular disease, suggesting conventional risk assessment may overlook behavioural contexts. We examined whether co-occurring lifestyle behaviours formed distinct phenotypes associated with SCA. METHODS:We analysed data from a prospective multicentre case-control study conducted in South Korea (2017-2023) including out-of-hospital cardiac arrest of presumed cardiac aetiology and community controls frequency-matched by age, sex, and residential area. Seven lifestyle indicators were assessed: smoking, physical activity, sleep duration, and intake of red meat, fish, fruit, and vegetables. Latent class analysis identified lifestyle phenotypes, and odds ratios (ORs) were estimated using a single fully adjusted logistic regression model informed by a directed acyclic graph. RESULTS:The primary analysis included 1,466 cases and 2,619 controls, identifying five phenotypes. The reference class (C2) showed the lowest observed proportion of SCA and a predominantly middle-frequency dietary pattern. Compared with C2, fully adjusted odds of SCA were higher for C1 (OR, 1.48; 95% confidence interval: 1.13-1.94) and C3 (1.95; 1.60-2.39), and markedly higher for C4 (8.93; 6.75-11.81) and C5 (15.20; 11.19-20.64). C5 was characterised by high reported intake frequency across food groups, particularly red meat and fish; C4 combined current smoking, physical inactivity, and inadequate sleep with low reported intake frequency across food groups. CONCLUSION:Cluster-derived lifestyle phenotypes were associated with differing odds of SCA. Although causal effects and absolute risk cannot be inferred from this case-control design, these multidomain patterns may help characterise behavioural profiles associated with SCA and inform future prevention research.
AIMS:The long-term depressive symptoms trajectories across the transition to diagnosed cardiovascular disease (CVD) remain poorly characterized. We aimed to investigate trajectories of depressive symptoms before and after incident CVD in multinational aging populations. METHODS AND RESULTS:In this multicohort study, we analyzed longitudinal data from the Health and Retirement Study (HRS, 2002-2018), English Longitudinal Study of Ageing (ELSA, 2002-2019), and China Health and Retirement Longitudinal Study (CHARLS, 2011-2018). Depressive symptoms were measured biennially by the 8-item (HRS and ELSA) and 10-item (CHARLS) Centre for Epidemiological Studies Depression scale and standardized to Z-scores. Incident CVD was ascertained by medical history. Data were analyzed using discontinuous growth curve modeling and pooled using meta-analysis. Among 28 704 included participants (mean [SD] age, 62.3 [9.8] years; 43.4% male), 7363 developed CVD during follow-up. Participants who developed CVD had higher levels of depressive symptoms than those without CVD (pooled β [95% confidence interval]: 0.161 SD [0.086-0.236]). A significant increase in depressive symptoms scores was observed around the time of CVD diagnosis (pooled β: 0.089 SD [0.060-0.118]). Additionally, depressive symptoms scores increased both in the years preceding CVD onset (pooled β: 0.007 SD/year [0.004-0.010]) and following diagnosis (pooled β: 0.009 SD/year [0.004-0.013]). CONCLUSION:Depressive symptoms show a gradual increase before CVD diagnosis, elevate further around the time of diagnosis, and continue to worsen thereafter. These findings call for the integration of mental health screening and management into CVD care at all stages, from primary prevention to long-term disease management.
BACKGROUND AND AIMS:The interaction between air pollutants and lifestyles on heart failure (HF) and its variation by age of HF onset remains unclear. We aim to assess the joint association of air pollutants and lifestyles with HF and examine whether it differs between early- and late-onset HF. METHODS:: A cohort of 233,341 UK Biobank participants (2006-2010 baseline, followed up until 2023) was analyzed. Air pollution (NO2, NOx, PM10, PM2.5, PM2.5-10) and lifestyle (physical activity, smoking, sleep, diet, drinking, BMI) scores were computed, where higher scores indicated exposure to heavier air pollution or more unhealthy lifestyles. Cox models evaluated their associations with HF risk, and interactions were tested multiplicatively and additively. RESULTS:: High air pollution (scores in top tertile) and unfavourable lifestyle (scores up to 5∼6) were associated with higher HF risk independently. Joint-exposure analyses revealed that unfavourable lifestyle and high air pollution were jointly linked to HF, which appeared stronger for early-onset HF [relative excess risk due to interaction: 0.44 (95%CI:0.14∼0.74) for all HF, 1.33 (95%CI:0.03∼2.62) for early-onset HF, 0.39 (95%CI:0.08∼0.70) for late-onset HF; p-value for the early-versus-late interaction contrast=0.08]. In turn, lifestyle differences [between unfavourable (5∼6) and favourable (0∼2)] were associated with larger HF incidence differences in high than low air pollution group, especially for early-onset HF [1.62-fold (95%CI:1.42∼1.84) for all HF, 3.35-fold (95%CI:1.76∼9.79) for early-onset HF, 1.52-fold (95%CI:1.33∼1.74) for late-onset HF]. CONCLUSIONS:: Air pollutants and lifestyles are jointly associated with HF risk, with young adults in air-polluted areas potentially gaining greater benefits from lifestyle differences.
