BACKGROUND:Effects of ω-3 fatty acids (FAs) supplementation on cardiovascular outcomes have been investigated in several randomized controlled trials (RCTs). The VITamin D and OmegA-3 TriaL (VITAL) is the largest trial that tested the effect of ω-3 FA supplementation (840 mg/d of eicosapentaenoic acid and docosahexaenoic acid in 1.2:1) in a primary prevention population in the United States, with nonsignificant results (P > 0.05) for major cardiovascular disease (CVD) events. OBJECTIVES:To reanalyze VITAL using Bayesian methods accounting for prior evidence. METHODS:The VITAL randomly assigned 25,871 older United States adults with a median follow-up of 5.3 y. On the basis of prior evidence from RCTs, we used Weibull proportional hazards models adopting the Hamiltonian Monte Carlo sampling method to estimate posterior hazard ratio (HR) for total CAD, myocardial infarction (MI), composite major CVD events (CAD/stroke/CVD death), CVD death, all-cause death, and stroke. Several distinct informative priors were formulated based on Bayesian hierarchical models of previous trials similar to VITAL. RESULTS:Bayesian analyses with the use of noninformative prior yielded essentially the same results as the corresponding frequentist analyses. The effects of ω-3 FA supplementation on CAD and MI were robust across the priors, with the posterior HR estimates varying from 0.88 to 0.93 and 0.82 to 0.90, respectively. Without skepticism into the priors, posterior HRs were 0.95-0.96 in CVD, 0.91-0.92 in cardiovascular death, and 0.95-0.96 in all-cause death risks, respectively. Stroke risk was unchanged by the intervention. According to primary informed priors, probabilities of ω-3 FA being effective were 99.7% for CAD, 99.6% for total MI, 98.4% for CVD, 98.8% for all-cause death, 99.8% for cardiovascular death, and 33.7% for stroke, respectively. CONCLUSIONS:Bayesian analyses of VITAL incorporating previous RCT evidence suggest that daily ω-3 FA supplementation robustly lowers risk of coronary events but not stroke, providing enhanced support for the primary prevention use of ω-3 FA supplementation for coronary events. TRIAL REGISTRATION NUMBER:This study was registered at VITAL clinicaltrials.gov identifier as NCT01169259.

Backgrounds: Cocoa flavanols have potential blood pressure (BP)-lowering effects in shorter-term, smaller-scale randomized clinical trials (RCTs), but its effect on incident hypertension has not been examined in a large-scale RCT. The COcoa Supplement and Multivitamin Outcomes Study (COSMOS) evaluated whether longer-term cocoa extract supplementation had an effect on incident hypertension in older adults. Methods: COSMOS is a 2x2 factorial, double-blind, placebo-controlled RCT testing cocoa extract (including 500 mg/day cocoa flavanols, with 80 mg/day (-)-epicatechin) and a multivitamin among 21,442 women aged ≥65y and men aged ≥60y. In 8,905 COSMOS participants free from baseline self-reported hypertension, we investigated the effect of cocoa extract on incident hypertension using Cox proportional hazards models. Incident hypertension was defined as self-reported first-time physician diagnosis, initiation of anti-hypertensive medications, or self-reported elevated systolic BP (SBP) ≥ 140 mmHg or diastolic BP (DBP) ≥ 90 mmHg. Results: Mean age at baseline was 71.1 years (SD: 6.2) and 58.8% of the participants were women. 48.6% had baseline SBP < 120 mmHg, corresponding to the normal range. Baseline characteristics were balanced between the cocoa extract and placebo groups. Over a median follow-up of 3.4 years, in intention-to-treat analyses cocoa extract supplementation had no significant effect on incident hypertension, with the incidence rates (IRs) of 71.2 and 73.8 per 1000 person-years in cocoa and placebo groups, respectively (hazard ratio [HR]: 0.96 [95% confidence interval: 0.88, 1.05]). A significant interaction was observed between treatment assignment and baseline SBP (p=0.002). Cocoa extract supplementation reduced the incidence of hypertension among participants with baseline SBP < 120 mmHg (IR: 38.2 vs. 50.3 per 1000 person-years; HR: 0.76 [0.64, 0.90]), but not among those with SBP of 120-139 mmHg (IR: 101.4 vs. 96.8 per 1000 person-years; HR: 1.05 [0.93, 1.18]; Figure ). Consistent results were obtained in sensitivity analyses using different definitions of hypertension: physician diagnosis, initiation of anti-hypertensive medications, or elevated BP alone. Per-protocol analysis using inverse probability weighting to account for non-adherence also supported the effect in SBP < 120 mmHg. Conclusion: In older adults, years-long supplementation of cocoa extract may reduce the risk of incident hypertension among those with normal BP levels.

BACKGROUND:Cocoa flavanols have potential blood pressure (BP)-lowering effects in shorter-term, smaller-scale randomized clinical trials, but their effect on incident hypertension has not been examined in a large-scale and long-term randomized clinical trial. METHODS:The COSMOS (Cocoa Supplement and Multivitamin Outcomes Study) is a 2×2 factorial, double-blind, placebo-controlled randomized clinical trial testing cocoa extract (including 500 mg/d cocoa flavanols, with 80 mg/d [-]-epicatechin) and a multivitamin among 21 442 women aged ≥65 years and men aged ≥60 years. Placebos did not include any bioactive compounds. In 8905 COSMOS participants free from baseline hypertension, we investigated the effect of cocoa extract on incident hypertension using Cox proportional hazards models. Incident hypertension was defined as self-reported first-time physician diagnosis, initiation of antihypertensive medications, or elevated BP. RESULTS:Mean age at baseline was 71.1 years (SD, 6.2), and 59% were women. Over a median follow-up of 3.4 years, cocoa extract supplementation had no significant effect on incident hypertension in an intention-to-treat analysis, with incidence rates of 7.1 and 7.4 per 100 person-years in cocoa and placebo groups, respectively (hazard ratio, 0.96 [95% CI, 0.88-1.05]). In subgroup analyses, cocoa extract supplementation reduced the incidence of hypertension among participants with baseline systolic BP <120 mm Hg (hazard ratio, 0.76 [0.64-0.90]), but not among those with systolic BP of 120 to 139 mm Hg (hazard ratio, 1.05 [0.93-1.18]; P-interaction=0.002). The effect among baseline systolic BP <120 mm Hg became evident at year 2 after randomization. CONCLUSIONS:In older adults, long-term cocoa extract supplementation did not reduce the overall risk of self-reported incident hypertension. However, among those with normal systolic BP at baseline, cocoa extract reduced hypertension risk by 24%.

Recent advancements in machine learning (ML) for analyzing heterogeneous treatment effects (HTE) are gaining prominence within the medical and epidemiological communities, offering potential breakthroughs in the realm of precision medicine by enabling the prediction of individual responses to treatments. This paper introduces the methodological frameworks used to study HTEs, particularly based on a single randomized controlled trial (RCT). We focus on methods to estimate conditional average treatment effect (CATE) for multiple covariates, aiming to predict individualized treatment effects. We explore a range of methodologies from basic frameworks like the T-learner, S-learner, and Causal Forest, to more advanced ones such as the DR-learner and R-learner, as well as cross-validation for CATE estimation to enhance statistical efficiency by estimating CATE for all RCT participants. We also provide a practical application of these approaches using the Preventing Overweight Using Novel Dietary Strategies (POUNDS Lost) trial, which compared the effects of high versus low-fat diet interventions on 2-year weight changes. We compared different sets of covariates for CATE estimation, showing that the DR- and R-learners are useful for the estimation of CATE in high-dimensional settings. This paper aims to explain the theoretical underpinnings and methodological nuances of ML-based HTE analysis without relying on technical jargon, making these concepts more accessible to the clinical and epidemiological research communities.

OBJECTIVE:To examine the long-term effect of cocoa flavanols on inflammaging biomarkers in the COcoa Supplement and Multivitamin Outcomes Study (COSMOS). METHODS:COSMOS is a large, randomised, double-blind, placebo-controlled, 2 × 2 factorial trial testing the effects of a cocoa extract supplement (containing 500 mg cocoa flavanols/day, including 80 mg (-)-epicatechin) among women aged ≥65 years and men aged ≥60 years. This ancillary study measured five widely used serum inflammaging biomarkers, including three pro-inflammatory markers (high-sensitivity C-reactive protein [hsCRP], interleukin-6, tumour necrosis factor-α), one anti-inflammatory cytokine (interleukin-10) and one pleotropic cytokine (interferon-γ [IFN-γ]) in a random sample of 598 participants with biospecimens collected at baseline, Year 1, and Year 2. RESULTS:The mean age was 70.0 ± 5.6 years, and 49.8% were female. Cocoa extract supplementation significantly decreased hsCRP levels compared with placebo, with a between-group difference in yearly percentage change relative to baseline levels of -8.4% (95% CI, -14.1% to -2.3%; nominal P = .008; Holm-adjusted P value = .039). Moreover, cocoa extract increased IFN-γ with a 6.8% (95% CI, 1.5% to 12.2%, nominal P = .011; Holm-adjusted P value = .043) difference in yearly percentage change versus placebo. The effects of cocoa extract on other inflammatory markers were not significant (all adjusted P values >.05). CONCLUSION:Cocoa extract supplementation significantly decreased hsCRP, supporting a role in modulating the chronic inflammaging process as a potential mechanism underlying its cardio-protective effects, including a 27% reduction in cardiovascular disease death in the COSMOS trial. The biological effect of increased IFN-γ by cocoa extract warrants further exploration.

Introduction: Although some trials have reported that multivitamin-multimineral (MVM) supplements reduce age-related chronic conditions such as cancer, cataracts, and cognitive decline among older adults, whether MVM slows the aging process remains unknown. Hypothesis: We hypothesized that a daily MVM could reduce biological aging measured by DNA methylation after 2 years of follow-up in the COcoa Supplement and Multivitamin Outcomes Study (COSMOS). Methods: COSMOS is a large, randomized, double-blind, placebo-controlled, 2×2 factorial trial testing a daily MVM (Centrum Silver) and cocoa extract among women aged ≥65 y and men aged ≥60 y free of major cardiovascular disease and recently diagnosed cancer. We included 958 participants from the COSMOS Blood subcohort with biospecimens collected at baseline, year 1, and year 2. We calculated five epigenetic aging measures encompassing the first generation trained on chronological age (PCHorvath, PCHannum), the second generation trained on all-cause mortality (PCPhenoAge, PCGrimAge), and the third generation trained on pace of aging (DunedinPACE) on the Infinium Human Methylation EPIC+ Array. For the first two generations, we calculated age acceleration (Accel) by residualizing these clocks for chronological age. Linear mixed-effects models evaluated the treatment effects of a MVM on each measure. Results: The mean age was 70.2±5.6 years, and 482 (50.3%) were female. Compared with placebo, daily MVM use significantly reduced the second generation of age clocks, with a between-group difference in yearly change of -0.111 years (95% CI, -0.203 to -0.018; P=0.02) for PCGrimAge Accel and -0.209 years (-0.405 to -0.013; P=0.037) for PCPhenoAge Accel. The effect of daily MVM use was not significant for PCHorvath Accel, PCHorvath Accel, or DunedinPACE. When stratified by baseline epigenetic aging measures, the effect of MVM on PCGrimAge Accel was more pronounced among those with higher baseline PCGrimAge Accel (-0.201 [-0.334 to -0.068]) than their counterparts (-0.014 [-0.135 to 0.107]; P interaction=0.041). The effect was consistent when stratified by other baseline characteristics.. Conclusion: Our study provides the first evidence from a large-scale, long-term, randomized controlled trial that MVM supplementation potentially slows biological aging among older adults, as measured by the PhenoAge and GrimAge epigenetic clocks, with more pronounced effects among those with greater baseline accelerated biological aging.

Background Exposure to lipopolysaccharide (LPS), a potent proinflammatory glycolipid derived from gut microbiota, may be linked to the development of coronary heart disease (CHD). However, evidence from human studies is limited. Objectives We aimed to investigate prospective relationships between 2 plasma biomarkers of LPS exposures—LPS-binding protein (LBP) and soluble cluster of differentiation 14 (sCD14)—in relation to incident CHD among United States males and females. Methods A prospective nested 1:1 matched case-control study of CHD was conducted among participants in the Nurses’ Health Study II (NHSII) and Health Professionals Follow-up Study (HPFS). Plasma concentrations of LBP and sCD14 were measured in 496 HPFS male CHD case-control pairs and 212 NHSII female pairs. Results Among controls, plasma concentrations of LBP exhibited positive correlations with age, body mass index, and C-reactive protein (CRP) concentrations and an inverse correlation with high-density lipoprotein cholesterol concentrations. For sCD14, positive correlations with age and CRP were only observed in HPFS controls. Neither elevated LBP nor sCD14 concentrations were significantly associated with incident CHD in HPFS. In NHSII, higher sCD14 concentrations, but not LBP, were significantly associated with higher risk of CHD, with a risk ratio of 3.01 [95% confidence interval (CI): 1.28, 7.11] when comparing extreme quintiles. Collectively, CRP and the total cholesterol/ high-density lipoprotein cholesterol ratio explained 27.9% (95% CI: 7.1%, 66.1%; P = 0.01) of the positive association between sCD14 and CHD in NHSII females. These associations were not modified by physical activity, alcohol intake, body mass index, inflammation markers, family history of CHD, or the presence of hypertension, hyperlipidemia, or type-2 diabetes. Conclusion Higher concentrations of sCD14 may be associated with an increased risk of CHD in females, whereas LBP concentrations are not associated with CHD in either sex. These data do not support that LPS exposure in initially healthy individuals is a contributing CHD risk factor, although the potential sex difference should be explored further.

Backgrounds: Passive heating practices, such as hot water immersion and Finnish sauna bathing, have been associated with reduced cardiovascular disease (CVD) risk in observational studies. This systematic review and meta-analysis aimed to comprehensively evaluate the present literature about non-acute effects of passive heating interventions on cardiometabolic and vascular health in randomized controlled trials (RCTs). Methods and Results: A comprehensive search was conducted in PubMed, Embase, and Cochrane Central Register of Controlled Trials up to November 4, 2024. We included RCTs that assessed the effects of passive heating interventions of at least one week duration on cardiometabolic and vascular outcomes in adults. Twenty RCTs that tested interventions such as hot water bathing, saunas, hot yoga, and local heating were included in the present review, with durations ranging from 2 to 15 weeks. Our analysis found no significant pooled effects for majority of outcomes, including flow-mediated dilation, pulse wave velocity, resting heart rate, heart rate variability, fasting glucose, HbA1c, total/HDL/LDL cholesterol, triglycerides, and C-reactive protein. For systolic blood pressure (SBP), the overall pooled estimate was not statistically significant (–2.46 mmHg [95 % CI: 5.02 to 0.10]; I²=60.3 %). However, subgroup analyses indicated a significant SBP reduction was confined to systemic (whole-body) heating interventions (–4.11 mmHg [–7.36, –0.86]) and in populations with underlying coronary risk or CVD (–2.52 mmHg [–4.26, –0.79]). Conclusions: Current evidence from RCTs indicates that passive heating interventions may not improve most of the cardiometabolic or vascular health markers. While a potential reduction in SBP was observed with systemic heating and among adults with coronary risk, this finding should be interpreted with caution due to high heterogeneity and limitations in the included studies.This systematic review is registered at PROSPERO (registration number: CRD42024621600).

Randomized controlled trials (RCTs) have demonstrated benefits of marine omega-3 polyunsaturated fatty acids (omega-3 FA) supplementation for the prevention of coronary heart disease (CHD). However, it has not been clear which individuals benefit the most from supplementation. We sought to develop an omega-3 effect score to stratify individuals according to their expected benefit from supplementation. Among the 25,871 randomized participants without a history of cardiovascular disease in the VITamin D and OmegA-3 TriaL (VITAL), we applied machine-learning (ML) approaches to predict individual treatment effect of omega-3 FA supplementation on 5-year CHD risk using 11 covariates pre-specified in the VITAL protocol. An omega-3 effect score was developed such that each covariate contributed linearly. ML algorithms effectively stratified participants by their expected benefit according to individual factors; for example, there was 1.21

Wearable devices provide an opportunity to remotely collect objective data on a patient’s physical activity (PA) and sleep patterns, blood pressure (BP), blood glucose, and other health-related metrics. To date, the prevalence of use and role of wearable devices in older adults diagnosed with cancer remains unclear. We sought to evaluate patterns and characteristics associated with wearable device use. A cross-sectional survey on wearable device use was nested within the COSMOS trial, a completed randomized trial of cocoa extract and multivitamin supplementation among U.S. older adults that began in 2015. A 2023 follow-up survey included questions about the usage and frequency of use of any wearable device (activity tracker, BP monitor, or glucose monitor), and willingness to share data for future research. We included participants who completed the 2023 survey and had a self-reported history of cancer diagnosis at baseline, except for non-melanoma skin cancers and compared survey responses to individuals who did not report a history of cancer. Multivariable logistic regression analyses were conducted to assess characteristics associated with wearable device usage adjusted for age, sex, race, income, body mass index (BMI), PA level, and diet quality. Of the 18, 552 COSMOS participants with 2023 survey data, 3, 045 reported a cancer history, 62% were female with a median age 73 years (interquartile range=8.3). There were 780 (26%), 1161 (38%), and 60 (2%) participants with a cancer history reporting usage of an activity tracker, BP monitor, or glucose monitor, respectively. Activity tracker usage was lower in those with vs without a cancer history (n=15, 507) (26% vs 29%, p=0.0003), while BP monitor usage was higher (38% vs 36%, p=0.023). Among those with a cancer history and reporting any wearable device usage, 78% used activity trackers, 19% used BP monitors, and 66% used glucose monitors most or all of the time. The percentage of individuals willing to share their data was higher for those with a history of cancer compared to those without, for all devices, with statistically significant differences for activity trackers (84% vs. 81%, p=0.02). In multivariable logistic regression, age (Odds ratio (OR): 1.04 per 1-year decrease), female sex (OR: 1.55), higher income (OR: 2.03), BMI (OR: 1.02), and PA level (OR: 1.07) were each statistically significantly (p<0.05) associated with increased odds of activity tracker use. BMI also was significantly associated with use of BP and glucose monitors (p<0.05). Older adults with a history of cancer have a high prevalence of wearable device usage, and are willing to share their data for research purposes. This highlights the potential to leverage existing trials to recruit participants in future studies involving wearables and collect valuable insights on their daily activity and biometric data. Gillian Gresham, Cami N. Christopher, Rikuta Hamaya, Aditi Hazra, JoAnn E. Manson, Howard D. Sesso. Use of wearable technology for physical activity and health monitoring in individuals diangnosed with cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 7365.

BACKGROUND:Increased blood pressure (BP) variability is linked to dementia risk, but the relationship between baroreflex sensitivity (BRS), a fundamental mechanism for maintaining stable BP, and dementia risk is undetermined. METHODS:We tested the hypothesis that impaired BRS is associated with increased dementia risk in 1819 older adults (63% women; age, 71.0±6.3 years) from the community-based Rotterdam Study. Cardiac BRS was determined from a 5-minute beat-to-beat BP recording at supine rest between 1997 and 1999. Cardiac BRS measures the correlation between changes in consecutive beat-to-beat systolic BP and subsequent responses in heartbeat intervals, with a higher value indicating better BRS. The primary outcome was incident dementia ascertained from baseline through January 1, 2020; the secondary outcome was all-cause mortality. RESULTS:During a median follow-up of 14.8 years, 421 participants developed dementia. The association of cardiac BRS with dementia risk differed by antihypertensive medication use (Pinteraction=0.03) and was only observed in participants not taking antihypertensives. Specifically, in those not taking antihypertensive medication, reduced BRS was associated with a higher risk of dementia (adjusted hazard ratio comparing bottom versus top quintiles, 1.60 [95% CI, 1.07-2.40]; Ptrend=0.02). Reduced BRS was also associated with an increased risk of death (corresponding hazard ratio, 1.76 [95% CI, 1.32-2.35]). The association remained after adjusting for average BP and BP variability. CONCLUSIONS:Impaired BRS partly explains hypertension-related brain damage and excessive dementia risk beyond conventional BP measures, making it a potential novel biomarker for the early detection and prevention of dementia.

Background Accurate quantification of sodium intake based on self‐reported dietary assessments has been a persistent challenge. We aimed to apply machine‐learning (ML) algorithms to predict 24‐hour urinary sodium excretion from self‐reported questionnaire information. Methods and Results We analyzed 3454 participants from the NHS (Nurses' Health Study), NHS‐II (Nurses' Health Study II), and HPFS (Health Professionals Follow‐Up Study), with repeated measures of 24‐hour urinary sodium excretion over 1 year. We used an ensemble approach to predict averaged 24‐hour urinary sodium excretion using 36 characteristics. The TOHP‐I (Trial of Hypertension Prevention I) was used for the external validation. The final ML algorithms were applied to 167 920 nonhypertensive adults with 30‐year follow‐up to estimate confounder‐adjusted hazard ratio (HR) of incident hypertension for predicted sodium. Averaged 24‐hour urinary sodium excretion was better predicted and calibrated with ML compared with the food frequency questionnaire (Spearman correlation coefficient, 0.51 [95% CI, 0.49–0.54] with ML; 0.19 [95% CI, 0.16–0.23] with the food frequency questionnaire; 0.46 [95% CI, 0.42–0.50] in the TOHP‐I). However, the prediction heavily depended on body size, and the prediction of energy‐adjusted 24‐hour sodium excretion was modestly better using ML. ML‐predicted sodium was modestly more strongly associated than food frequency questionnaire‐based sodium in the NHS‐II (HR comparing Q5 versus Q1, 1.48 [95% CI, 1.40–1.56] with ML; 1.04 [95% CI, 0.99–1.08] with the food frequency questionnaire), but no material differences were observed in the NHS or HPFS. Conclusions The present ML algorithm improved prediction of participants' absolute 24‐hour urinary sodium excretion. The present algorithms may be a generalizable approach for predicting absolute sodium intake but do not substantially reduce the bias stemming from measurement error in disease associations.

Background: Evidence is lacking on the relative contributions of specific lifestyle factors and their overall contribution to prevention of hypertension, in particular early-onset hypertension. Methods: This prospective cohort study included participants of the Nurses' Health Study (NHS, N = 52,780 women, aged 40-67 in 1986), the NHS II (N = 83,871 women, aged 27-46 in 1991), and the Health Professionals Follow-up Study (HPFS, N = 31,269 men, aged 40-75 in 1986), who were free from hypertension, cardiovascular disease and cancer at baseline. Four modifiable lifestyles were evaluated based on hypertension guidelines: BMI, moderate-to-vigorous physical activity, Dietary Approaches to Stop Hypertension (DASH) score, and alcohol intake. Primary outcome was incident self-reported diagnosis of hypertension with 27-31 years of follow-up. Results: Each lifestyle factor was associated with incident hypertension in dose-dependent manners across the cohorts, with BMI having the strongest associations. On average, adhering to BMI <25 kg/m(2) was associated with 20.3 [18.5, 22.0], 25.0 [23.2, 26.8], and 18.6 [16.7, 20.7] months longer periods free from hypertension during 25-year follow-up in each cohort respectively. BMI accounted for approximately 20 % of incident hypertension in NHS and HPFS, and 35 % of early-onset hypertension (age < 55 y). Moderate-to-vigorous physical activity and diet accounted for 10-15 % of incident hypertension in women, and the contributions were greater for early-onset hypertension. Conclusion: Healthy weight during adulthood was most substantially associated with incident hypertension among lifestyle factors, but diet, physical activity, and alcohol intake were also related to the risk across all ages, and hypertension-free periods, with stronger associations in early-onset hypertension.

Introduction: With advancement of medicine, alternative exposures or interventions are emerging with respect to a common outcome, and there are needs to formally test the difference in the associations of multiple exposures. Methods: The paper proposes a duplication method-based multivariate Wald test in the Cox proportional hazard regression model to test the difference in the associations of multiple exposures with a same outcome. This method applies to continuous or categorical exposures. For illustration, the method was applied to compare the associations between alignment to 2 different dietary patterns (Alternative Healthy Eating Index-2010 and reversed empirical dietary inflammatory pattern), either as continuous or quartile exposures, and incident chronic diseases, defined as a composite of cardiovascular disease, cancer, and diabetes, in the Health Professional Follow-up Study. Relevant sample codes in R that implement the proposed approach are provided. The present analysis was conducted in 2024. Results: With a median follow-up of 22 years, there were a total of 14,427 chronic disease incidences among N=43,185 men included in the Health Professional Follow-up Study analysis. The hazard ratios (HRs) (95% CI) per increment from the 10th to the 90th percentile for incident chronic disease were 0.83 (0.79, 0.87) for Alternative Healthy Eating Index-2010 and 0.76 (0.73, 0.80) for reversed empirical dietary inflammatory pattern. Although the 95% CIs were overlapped for each exposure, the proposed test was well powered to detect the difference (p=0.005). Conclusions: The proposed duplication-method-based approach offers a flexible, formal statistical test for heterogeneity in the associations of multiple exposures with the common outcome with minimal assumptions.

Importance:The associations between angiographic findings and post-percutaneous coronary intervention (PCI) fractional flow reserve (FFR) and their clinical relevance according to residual functional disease burden have not been thoroughly investigated. Objectives:To evaluate the association of angiographic and physiologic parameters according to residual functional disease burden after drug-eluting stent implantation. Design, Setting, and Participants:This cohort study population was from the International Post-PCI FFR registry, which incorporated 4 registries from Korea, China, and Japan. Patients who underwent angiographically successful second-generation drug-eluting stent implantation and post-PCI FFR measurement were included in the analysis. The patients were divided into 3 groups according to the residual disease burden (post-PCI FFR ≤0.80 [residual ischemia], 0.81-0.86 [suboptimal], and >0.86 [optimal]). The data were collected from August 23, 2018, to June 11, 2019, and the current analysis was performed from January 11, 2022, to October 7, 2023. Exposures:Angiographic parameters and post-PCI FFR. Main Outcomes and Measures:The primary outcome was target vessel failure (TVF), defined as a composite of cardiac death, target vessel-related myocardial infarction, and target vessel revascularization (TVR) at 2 years. Results:In this cohort of 2147 patients, the mean (SD) age was 64.3 (10.0) years, and 1644 patients (76.6%) were men. Based on the post-PCI physiologic status, 269 patients (12.5%) had residual ischemia, 551 (25.7%) had suboptimal results, and 1327 (61.8%) had optimal results. Angiographic parameters had poor correlations with post-PCI FFR (r < 0.20). Post-PCI FFR was isolated from all angiographic parameters in the unsupervised hierarchical cluster analysis. Post-PCI FFR was associated with the occurrence of TVF (adjusted hazard ratio [AHR] per post-PCI FFR 0.01 increase, 0.94 [95% CI, 0.92-0.97]; P < .001), but angiographic parameters were not. The residual ischemia group had a significantly higher rate of TVF than the suboptimal group (AHR, 1.75 [95% CI, 1.08-2.83]; P = .02) and the optimal group (AHR, 2.94 [95% CI, 1.82-4.73]; P < .001). The TVR in the residual ischemia group was predominantly associated with TVR in the nonstented segment (14 [53.8%]), unlike the other 2 groups (3 [10.0%] in the suboptimal group and 13 [30.2%] in the optimal group). Conclusions and Relevance:In this cohort study of the International Post-PCI FFR registry, a low degree of associations were observed between angiographic and physiologic parameters after PCI. Post-PCI FFR, unlike angiographic parameters, was associated with clinical events and the distribution of clinical events. The current study supports the use of post-PCI FFR as a procedural quality metric and further prospective study is warranted.

With advancement of medicine, alternative exposures or interventions are emerging with respect to a common outcome, and there are needs to formally test the difference in the associations of multiple exposures. We propose a duplication method-based multivariate Wald test in the Cox proportional hazard regression analyses to test the difference in the associations of multiple exposures with a same outcome. The proposed method applies to linear or categorical exposures. To illustrate our method, we applied our method to compare the associations between alignment to two different dietary patterns, either as continuous or quartile exposures, and incident chronic diseases, defined as a composite of CVD, cancer, and diabetes, in the Health Professional Follow-up Study. Relevant sample codes in R that implement the proposed approach are provided. The proposed duplication-method-based approach offers a flexible, formal statistical test of multiple exposures for the common outcome with minimal assumptions.

BACKGROUND:Several metabolites are individually related to incident type 2 diabetes (T2D) risk. We prospectively evaluated a novel T2D-metabolite pattern with a risk of progression to T2D among high-risk women with a history of gestational diabetes mellitus (GDM). METHODS:The longitudinal Nurses' Health Study II cohort enroled 116,429 women in 1989 and collected blood samples from 1996 to 1999. We profiled plasma metabolites in 175 incident T2D cases and 175 age-matched controls, all with a history of GDM before the blood draw. We derived a metabolomics score from 21 metabolites previously associated with incident T2D in the published literature by scoring according to the participants' quintile (1-5 points) of each metabolite. We modelled the T2D metabolomics score categorically in quartiles and continuously per 1 standard deviation (SD) with the risk of incident T2D using conditional logistic regression models adjusting for body mass index at the blood draw, and other established T2D risk factors. RESULTS:The percentage of women progressing to T2D ranged from 10% in the bottom T2D metabolomics score quartile to 78% in the highest score quartile. Adjusting for established T2D risk factors, women in the highest quartile had more than a 20-fold greater diabetes risk than women in the lowest quartile (odds ratios [OR] = 23.1 [95% CI = 8.6, 62.1]; p for trend<0.001). The continuous T2D metabolomics score was strongly and positively associated with incident T2D (adjusted OR = 2.7 per SD [95% CI = 1.9, 3.7], p < 0.0001). CONCLUSIONS:A pattern of plasma metabolites among high-risk women is associated with a markedly elevated risk of progression to T2D later in life.

Importance:Current US physical activity (PA) guidelines prescribe moderate to vigorous PA (MVPA) time of at least 150 minutes per week for health. An analogous step-based recommendation has not been issued due to insufficient evidence. Objective:To examine the associations of MVPA time and step counts with all-cause mortality and cardiovascular disease (CVD). Design, Setting, and Participants:This cohort study analyzed data from an ongoing follow-up study of surviving participants of the Women's Health Study, a randomized clinical trial conducted from 1992 to 2004 in the US to evaluate use of low-dose aspirin and vitamin E for preventing cancer and CVD. Participants were 62 years or older who were free from CVD and cancer, completed annual questionnaires, and agreed to measure their PA with an accelerometer as part of a 2011-2015 ancillary study. Participants were followed up through December 31, 2022. Exposures:Time spent in MVPA and step counts, measured with an accelerometer for 7 consecutive days. Main Outcomes and Measures:The associations of MVPA time and step counts with all-cause mortality and CVD (composite of myocardial infarction, stroke, and CVD mortality) adjusted for confounders. Cox proportional hazards regression models, restricted mean survival time differences, and area under the receiver operating characteristic curve (AUC) were used to evaluate the associations. Results:A total of 14 399 women (mean [SD] age, 71.8 [5.6] years) were included. The median (IQR) MVPA time and step counts were 62 (20-149) minutes per week and 5183 (3691-7001) steps per day, respectively. During a median (IQR) follow-up of 9.0 (8.0-9.9) years, the hazard ratios (HR) per SD for all-cause mortality were 0.82 (95% CI, 0.75-0.90) for MVPA time and 0.74 (95% CI, 0.69-0.80) for step counts. Greater MVPA time and step counts (top 3 quartiles vs bottom quartile) were associated with a longer period free from death: 2.22 (95% CI, 1.58-2.85) months and 2.36 (95% CI, 1.73-2.99) months at 9 years follow-up, respectively. The AUCs for all-cause mortality from MVPA time and step counts were similar: 0.55 (95% CI, 0.52-0.57) for both metrics. Similar associations of these 2 metrics with CVD were observed. Conclusion and Relevance:Results of this study suggest that among females 62 years or older, MVPA time and step counts were qualitatively similar in their associations with all-cause mortality and CVD. Step count-based goals should be considered for future guidelines along with time-based goals, allowing for the accommodation of personal preferences.