N-Glycosylation Profiles of Immunoglobulin G and Future Cardiovascular Events.

BACKGROUND: Posttranslational glycosylation of IgG can modulate its inflammatory capacity through structural variations. We examined the association of baseline IgG N-glycans and an IgG glycan score with incident cardiovascular disease (CVD). METHODS: IgG N-glycans were measured in 2 nested CVD case-control studies: JUPITER (Justification for the Use of Statins in Prevention: an Intervention Trial Evaluating Rosuvastatin; NCT00239681; primary prevention; discovery; Npairs=162); and TNT trial (Treating to New Targets; NCT00327691; secondary prevention; validation; Npairs=397). Using conditional logistic regression, we investigated the association of future CVD with baseline IgG N-glycans and a glycan score adjusting for clinical risk factors (statin treatment, age, sex, race, lipids, hypertension, and smoking) in JUPITER. Significant associations were validated in TNT, using a similar model further adjusted for diabetes. Using least absolute shrinkage and selection operator regression, an IgG glycan score was derived in JUPITER as a linear combination of selected IgG N-glycans. RESULTS: Six IgG N-glycans were associated with CVD in both studies: an agalactosylated glycan (IgG-GP4) was positively associated, while 3 digalactosylated glycans (IgG glycan peaks 12, 13, 14) and 2 monosialylated glycans (IgG glycan peaks 18, 20) were negatively associated with CVD after multiple testing correction (overall false discovery rate <0.05). Four selected IgG N-glycans comprised the IgG glycan score, which was associated with CVD in JUPITER (adjusted hazard ratio per glycan score SD, 2.08 [95% CI, 1.52-2.84]) and validated in TNT (adjusted hazard ratio per SD, 1.20 [95% CI, 1.03-1.39]). The area under the curve changed from 0.693 for the model without the score to 0.728 with the score in JUPITER (PLRT=1.1×10-6) and from 0.635 to 0.637 in TNT (PLRT=0.017). CONCLUSIONS: An IgG N-glycan profile was associated with incident CVD in 2 populations (primary and secondary prevention), involving an agalactosylated glycan associated with increased risk of CVD, while several digalactosylated and sialylated IgG glycans associated with decreased risk. An IgG glycan score was positively associated with future CVD.

One-Year Effects of High-Intensity Statin on Bioactive Lipids: Findings From the JUPITER Trial.

BACKGROUND: Statin effects extend beyond low-density lipoprotein cholesterol reduction, potentially modulating the metabolism of bioactive lipids (BALs), crucial for biological signaling and inflammation. These bioactive metabolites may serve as metabolic footprints, helping uncover underlying processes linked to pleiotropic effects of statins and yielding a better understanding of their cardioprotective properties. This study aimed to investigate the impact of high-intensity statin therapy versus placebo on plasma BALs in the JUPITER trial (Justification for the Use of Statins in Prevention: an Intervention Trial Evaluating Rosuvastatin; NCT00239681), a randomized primary prevention trial involving individuals with low-density lipoprotein cholesterol <130 mg/dL and high-sensitivity C-reactive protein ≥2 mg/L. METHODS: Using a nontargeted mass spectrometry approach, over 11 000 lipid features were assayed from baseline and 1-year plasma samples from cardiovascular disease noncases from 2 nonoverlapping nested substudies: JUPITERdiscovery (n=589) and JUPITERvalidation (n=409). The effect of randomized allocation of rosuvastatin 20 mg versus placebo on BALs was examined by fitting a linear regression with delta values (∆=year 1-baseline) adjusted for age and baseline levels of each feature. Significant associations in discovery were analyzed in the validation cohort. Multiple comparisons were adjusted using 2-stage overall false discovery rate. RESULTS: We identified 610 lipid features associated with statin randomization with significant replication (overall false discovery rate, <0.05), including 26 with annotations. Statin therapy significantly increased levels of 276 features, including BALs with anti-inflammatory activity and arterial vasodilation properties. Concurrently, 334 features were significantly lowered by statin therapy, including arachidonic acid and proinflammatory and proplatelet aggregation BALs. By contrast, statin therapy reduced an eicosapentaenoic acid-derived hydroxyeicosapentaenoic acid metabolite, which may be related to impaired glucose metabolism. Additionally, we observed sex-related differences in 6 lipid metabolites and 6 unknown features. CONCLUSIONS: Statin allocation was significantly associated with upregulation of BALs with anti-inflammatory, antiplatelet aggregation and antioxidant properties and downregulation of BALs with proinflammatory and proplatelet aggregation activity, supporting the pleiotropic effects of statins beyond low-density lipoprotein cholesterol reduction.

Plasma Biomarker Profiles for Premature and Nonpremature Coronary Heart Disease in Women.

BACKGROUND: Premature coronary heart disease (CHD) is a major cause of death in women. We aimed to characterize biomarker profiles of women who developed CHD before and after age 65 years. METHODS: In the Women's Health Study (median follow-up 21.5 years), women were grouped by age and timing of incident CHD: baseline age <65 years with premature CHD by age 65 years (25 042 women; 447 events) and baseline age ≥65 years with nonpremature CHD (2982 women; 351 events). Associations of 44 baseline plasma biomarkers measured using standard assays and a nuclear magnetic resonance (NMR)-metabolomics assay were analyzed using Cox models adjusted for clinical risk factors. RESULTS: Twelve biomarkers showed associations only with premature CHD and included lipoprotein(a), which was associated with premature CHD [adjusted hazard ratio (HR) per SD: 1.29 (95% CI 1.17-1.42)] but not with nonpremature CHD [1.09(0.98-1.22)](Pinteraction = 0.02). NMR-measured lipoprotein insulin resistance was associated with the highest risk of premature CHD [1.92 (1.52-2.42)] but was not associated with nonpremature CHD (Pinteraction <0.001). Eleven biomarkers showed stronger associations with premature vs nonpremature CHD, including apolipoprotein B. Nine NMR biomarkers showed no association with premature or nonpremature CHD, whereas 12 biomarkers showed similar significant associations with premature and nonpremature CHD, respectively, including low-density lipoprotein (LDL) cholesterol [1.30(1.20-1.45) and 1.22(1.10-1.35)] and C-reactive protein [1.34(1.19-1.50) and 1.25(1.08-1.44)]. CONCLUSIONS: In women, a profile of 12 biomarkers was selectively associated with premature CHD, driven by lipoprotein(a) and insulin-resistant atherogenic dyslipoproteinemia. This has implications for the development of biomarker panels to screen for premature CHD.

Association of Physical Activity With Bioactive Lipids and Cardiovascular Events.

BACKGROUND: To clarify the mechanisms underlying physical activity (PA)-related cardioprotection, we examined the association of PA with plasma bioactive lipids (BALs) and cardiovascular disease (CVD) events. We additionally performed genome-wide associations. METHODS: PA-bioactive lipid associations were examined in VITAL (VITamin D and OmegA-3 TriaL)-clinical translational science center (REGISTRATION: URL: https://www. CLINICALTRIALS: gov; Unique identifier: NCT01169259; N=1032) and validated in JUPITER (Justification for the Use of statins in Prevention: an Intervention Trial Evaluating Rosuvastatin)-NC (REGISTRATION: URL: https://www. CLINICALTRIALS: gov; Unique identifier: NCT00239681; N=589), using linear models adjusted for age, sex, race, low-density lipoprotein-cholesterol, total-C, and smoking. Significant BALs were carried over to examine associations with incident CVD in 2 nested CVD case-control studies: VITAL-CVD (741 case-control pairs) and JUPITER-CVD (415 case-control pairs; validation). RESULTS: We detected 145 PA-bioactive lipid validated associations (false discovery rate <0.1). Annotations were found for 6 of these BALs: 12,13-diHOME, 9,10-diHOME, lysoPC(15:0), oxymorphone-3b-D-glucuronide, cortisone, and oleoyl-glycerol. Genetic analysis within JUPITER-NC showed associations of 32 PA-related BALs with 22 single-nucleotide polymorphisms. From PA-related BALs, 12 are associated with CVD. CONCLUSIONS: We identified a PA-related bioactive lipidome profile out of which 12 BALs also had opposite associations with incident CVD events.

Pleiotropy-Based Decomposition of Genetic Risk Scores: Association and Interaction Analysis for Type 2 Diabetes and CAD.

Genetic risk for a disease in the population may be represented as a genetic risk score (GRS) constructed as the sum of inherited risk alleles, weighted by allelic effects established in an independent population. While this formulation captures overall genetic risk, it typically does not address risk due to specific biological mechanisms or pathways that may nevertheless be important for interpretation or treatment response. Here, a GRS for disease is resolved into independent or nearly independent components pertaining to biological mechanisms inferred from pleiotropic relationships. The component GRSs' weights are derived from the singular value decomposition (SVD) of the matrix of appropriately scaled genetic effects, i.e., beta coefficients, of the disease variants across a panel of the disease-related phenotypes. The SVD-based formalism also associates combinations of disease-related phenotypes with inferred disease pathways. Applied to incident type 2 diabetes (T2D) in the Women's Genome Health Study (N = 23,294), component GRSs discriminate glycemic control and lipid-based genetic risk, while revealing significant interactions between specific components and BMI or physical activity, the latter not observed with a GRS for overall T2D genetic liability. Applied to coronary artery disease (CAD) in both the WGHS and in JUPITER (N = 8,749), a randomized trial of rosuvastatin for primary prevention of CVD, component GRSs discriminate genetic risk associated with LDL-C from risk associated with reciprocal genetic effects on triglycerides and HDL-C. They also inform the pharmacogenetics of statin treatment by demonstrating that benefit from rosuvastatin is as strongly related to genetic risk from triglycerides and HDL-C as from LDL-C.

Measures for evaluation of prognostic improvement under multivariate normality for nested and nonnested models.

When comparing performances of two risk prediction models, several metrics exist to quantify prognostic improvement, including the change in the area under the Receiver Operating Characteristic curve, the Integrated Discrimination Improvement, the Net Reclassification Index at event rate, the change in Standardized Net Benefit, the change in Brier score, and the change in scaled Brier score. We explore the behavior and interrelationships between these metrics under multivariate normality in nested and nonnested model comparisons. We demonstrate that, within the framework of linear discriminant analysis, all six statistics are functions of squared Mahalanobis distance, a robust metric that properly measures discrimination by quantifying the separation between the risk scores of events and nonevents. These relationships are important for overall interpretability and clinical usefulness. Through simulation, we demonstrate that the performance of the theoretical estimators under normality is comparable or superior to empirical estimation methods typically used by investigators. In particular, the theoretical estimators for the Net Reclassification Index and the change in Standardized Net Benefit exhibit less variability in their estimates as compared to their empirically estimated counterparts. Finally, we explore how these metrics behave with potentially nonnormal data by applying these methods in a practical example based on the sex-specific cardiovascular disease risk models from the Framingham Heart Study. Our findings aim to give greater insight into the behavior of these measures and the connections existing among them and to provide additional estimation methods with less variability for the Net Reclassification Index and the change in Standardized Net Benefit.

Cholesterol Efflux Capacity, High-Density Lipoprotein Particle Number, and Incident Cardiovascular Events: An Analysis From the JUPITER Trial (Justification for the Use of Statins in Prevention: An Intervention Trial Evaluating Rosuvastatin).

BACKGROUND: Recent failures of drugs that raised high-density lipoprotein (HDL) cholesterol levels to reduce cardiovascular events in clinical trials have led to increased interest in alternative indices of HDL quality, such as cholesterol efflux capacity, and HDL quantity, such as HDL particle number. However, no studies have directly compared these metrics in a contemporary population that includes potent statin therapy and low low-density lipoprotein cholesterol. METHODS: HDL cholesterol levels, apolipoprotein A-I, cholesterol efflux capacity, and HDL particle number were assessed at baseline and 12 months in a nested case-control study of the JUPITER trial (Justification for the Use of Statins in Prevention: An Intervention Trial Evaluating Rosuvastatin), a randomized primary prevention trial that compared rosuvastatin treatment to placebo in individuals with normal low-density lipoprotein cholesterol but increased C-reactive protein levels. In total, 314 cases of incident cardiovascular disease (CVD) (myocardial infarction, unstable angina, arterial revascularization, stroke, or cardiovascular death) were compared to age- and gender-matched controls. Conditional logistic regression models adjusting for risk factors evaluated associations between HDL-related biomarkers and incident CVD. RESULTS: Cholesterol efflux capacity was moderately correlated with HDL cholesterol, apolipoprotein A-I, and HDL particle number (Spearman r= 0.39, 0.48, and 0.39 respectively; P<0.001). Baseline HDL particle number was inversely associated with incident CVD (adjusted odds ratio per SD increment [OR/SD], 0.69; 95% confidence interval [CI], 0.56-0.86; P<0.001), whereas no significant association was found for baseline cholesterol efflux capacity (OR/SD, 0.89; 95% CI, 0.72-1.10; P=0.28), HDL cholesterol (OR/SD, 0.82; 95% CI, 0.66-1.02; P=0.08), or apolipoprotein A-I (OR/SD, 0.83; 95% CI, 0.67-1.03; P=0.08). Twelve months of rosuvastatin (20 mg/day) did not change cholesterol efflux capacity (average percentage change -1.5%, 95% CI, -13.3 to +10.2; P=0.80), but increased HDL cholesterol (+7.7%), apolipoprotein A-I (+4.3%), and HDL particle number (+5.2%). On-statin cholesterol efflux capacity was inversely associated with incident CVD (OR/SD, 0.62; 95% CI, 0.42-0.92; P=0.02), although HDL particle number again emerged as the strongest predictor (OR/SD, 0.51; 95% CI, 0.33-0.77; P<0.001). CONCLUSIONS: In JUPITER, cholesterol efflux capacity was associated with incident CVD in individuals on potent statin therapy but not at baseline. For both baseline and on-statin analyses, HDL particle number was the strongest of 4 HDL-related biomarkers as an inverse predictor of incident events and biomarker of residual risk. CLINICAL TRIAL REGISTRATION: URL: http://www.clinicaltrials.gov. Unique identifier: NCT00239681.

Clinical risk reclassification at 10 years.

Three papers in this issue focus on the role of calibration in model fit statistics, including the net reclassification improvement (NRI) and integrated discrimination improvement (IDI). This commentary reviews the development of such reclassification statistics along with more recent advances in our understanding of these measures. We show how the two-category NRI and the IDI are affected by changes in the event rate in theory and in an applied example. We also describe the role of calibration and how it may be assessed. Finally, we discuss the relevance of the event rate NRI for clinical use. Copyright © 2017 John Wiley & Sons, Ltd.

Asymptotic distribution of ∆AUC, NRIs, and IDI based on theory of U-statistics.

The change in area under the curve (∆AUC), the integrated discrimination improvement (IDI), and net reclassification index (NRI) are commonly used measures of risk prediction model performance. Some authors have reported good validity of associated methods of estimating their standard errors (SE) and construction of confidence intervals, whereas others have questioned their performance. To address these issues, we unite the ∆AUC, IDI, and three versions of the NRI under the umbrella of the U-statistics family. We rigorously show that the asymptotic behavior of ∆AUC, NRIs, and IDI fits the asymptotic distribution theory developed for U-statistics. We prove that the ∆AUC, NRIs, and IDI are asymptotically normal, unless they compare nested models under the null hypothesis. In the latter case, asymptotic normality and existing SE estimates cannot be applied to ∆AUC, NRIs, or IDI. In the former case, SE formulas proposed in the literature are equivalent to SE formulas obtained from U-statistics theory if we ignore adjustment for estimated parameters. We use Sukhatme-Randles-deWet condition to determine when adjustment for estimated parameters is necessary. We show that adjustment is not necessary for SEs of the ∆AUC and two versions of the NRI when added predictor variables are significant and normally distributed. The SEs of the IDI and three-category NRI should always be adjusted for estimated parameters. These results allow us to define when existing formulas for SE estimates can be used and when resampling methods such as the bootstrap should be used instead when comparing nested models. We also use the U-statistic theory to develop a new SE estimate of ∆AUC. Copyright © 2017 John Wiley & Sons, Ltd.

Tests of calibration and goodness-of-fit in the survival setting.

To access the calibration of a predictive model in a survival analysis setting, several authors have extended the Hosmer-Lemeshow goodness-of-fit test to survival data. Grønnesby and Borgan developed a test under the proportional hazards assumption, and Nam and D'Agostino developed a nonparametric test that is applicable in a more general survival setting for data with limited censoring. We analyze the performance of the two tests and show that the Grønnesby-Borgan test attains appropriate size in a variety of settings, whereas the Nam-D'Agostino method has a higher than nominal Type 1 error when there is more than trivial censoring. Both tests are sensitive to small cell sizes. We develop a modification of the Nam-D'Agostino test to allow for higher censoring rates. We show that this modified Nam-D'Agostino test has appropriate control of Type 1 error and comparable power to the Grønnesby-Borgan test and is applicable to settings other than proportional hazards. We also discuss the application to small cell sizes.

Mediterranean Diet Adherence and Risk of All-Cause Mortality in Women.

Importance: Higher adherence to the Mediterranean diet has been associated with reduced risk of all-cause mortality, but data on underlying molecular mechanisms over long follow-up are limited. Objectives: To investigate Mediterranean diet adherence and risk of all-cause mortality and to examine the relative contribution of cardiometabolic factors to this risk reduction. Design, Setting, and Participants: This cohort study included initially healthy women from the Women's Health Study, who had provided blood samples, biomarker measurements, and dietary information. Baseline data included self-reported demographics and a validated food-frequency questionnaire. The data collection period was from April 1993 to January 1996, and data analysis took place from June 2018 to November 2023. Exposures: Mediterranean diet score (range, 0-9) was computed based on 9 dietary components. Main Outcome and Measures: Thirty-three blood biomarkers, including traditional and novel lipid, lipoprotein, apolipoprotein, inflammation, insulin resistance, and metabolism measurements, were evaluated at baseline using standard assays and nuclear magnetic resonance spectroscopy. Mortality and cause of death were determined from medical and death records. Cox proportional hazards regression was used to calculate hazard ratios (HRs) for Mediterranean diet adherence and mortality risk, and mediation analyses were used to calculate the mediated effect of different biomarkers in understanding this association. Results: Among 25 315 participants, the mean (SD) baseline age was 54.6 (7.1) years, with 329 (1.3%) Asian women, 406 (1.6%) Black women, 240 (0.9%) Hispanic women, 24 036 (94.9%) White women, and 95 (0.4%) women with other race and ethnicity; the median (IQR) Mediterranean diet adherence score was 4.0 (3.0-5.0). Over a mean (SD) of 24.7 (4.8) years of follow-up, 3879 deaths occurred. Compared with low Mediterranean diet adherence (score 0-3), adjusted risk reductions were observed for middle (score 4-5) and upper (score 6-9) groups, with HRs of 0.84 (95% CI, 0.78-0.90) and 0.77 (95% CI, 0.70-0.84), respectively (P for trend < .001). Further adjusting for lifestyle factors attenuated the risk reductions, but they remained statistically significant (middle adherence group: HR, 0.92 [95% CI, 0.85-0.99]; upper adherence group: HR, 0.89 [95% CI, 0.82-0.98]; P for trend = .001). Of the biomarkers examined, small molecule metabolites and inflammatory biomarkers contributed most to the lower mortality risk (explaining 14.8% and 13.0%, respectively, of the association), followed by triglyceride-rich lipoproteins (10.2%), body mass index (10.2%), and insulin resistance (7.4%). Other pathways, including branched-chain amino acids, high-density lipoproteins, low-density lipoproteins, glycemic measures, and hypertension, had smaller contributions (<3%). Conclusions and Relevance: In this cohort study, higher adherence to the Mediterranean diet was associated with 23% lower risk of all-cause mortality. This inverse association was partially explained by multiple cardiometabolic factors.

Impact of correlation on predictive ability of biomarkers.

In this paper, we investigate how the correlation structure of independent variables affects the discrimination of risk prediction model. Using multivariate normal data and binary outcome, we prove that zero correlation among predictors is often detrimental for discrimination in a risk prediction model and negatively correlated predictors with positive effect sizes are beneficial. A very high multiple R-squared from regressing the new predictor on the old ones can also be beneficial. As a practical guide to new variable selection, we recommend to select predictors that have negative correlation with the risk score based on the existing variables. This step is easy to implement even when the number of new predictors is large. We illustrate our results by using real-life Framingham data suggesting that the conclusions hold outside of normality. The findings presented in this paper might be useful for preliminary selection of potentially important predictors, especially is situations where the number of predictors is large.

The Mediterranean Diet, Cardiometabolic Biomarkers, and Risk of All-Cause Mortality: A 25-Year Follow-Up Study of the Women's Health Study.

Background: Higher consumption of Mediterranean diet (MED) intake has been associated with reduced risk of all-cause mortality but limited data are available examining long-term outcomes in women or the underlying molecular mechanisms of this inverse association in human populations. We aimed to investigate the association of MED intake with long-term risk of all-cause mortality in women and to better characterize the relative contribution of traditional and novel cardiometabolic factors to the MED-related risk reduction in morality. Methods: In a prospective cohort study of 25,315 initially healthy women from the Women's Health Study, we assessed dietary MED intake using a validated semiquantitative food frequency questionnaire according to the usual 9-category measure of MED adherence. Baseline levels of more than thirty cardiometabolic biomarkers were measured using standard assays and targeted nuclear magnetic resonance spectroscopy, including lipids, lipoproteins, apolipoproteins, inflammation, glucose metabolism and insulin resistance, branched-chain amino acids, small metabolites, and clinical factors. Mortality and cause of death was ascertained prospectively through medical and death records. Results: During a mean follow-up of 25 years, 3,879 deaths were ascertained. Compared to the reference group of low MED intake (0-3, approximately the bottom tertile), and adjusting for age, treatment, and energy intake, risk reductions were observed for the middle and upper MED groups with respective HRs of 0.84 (95% CI 0.78-0.90) and 0.77 (95% CI 0.70-0.84), p for trend <0.0001. Further adjusting for smoking, physical activity, alcohol intake and menopausal factors attenuated the risk reductions which remained significant with respective HRs of 0.92 (95% CI 0.85-0.99) and 0.89 (95% CI 0.82-0.98), p for trend 0.0011. Risk reductions were generally similar for CVD and non-CVD mortality. Small molecule metabolites (e.g., alanine and homocysteine) and inflammation made the largest contributions to lower mortality risk (accounting for 14.8% and 13.0% of the benefit of the MED-mortality association, respectively), followed by triglyceride-rich lipoproteins (10.2%), adiposity (10.2%) and insulin resistance (7.4%), with lesser contributions (<3%) from other pathways including branched-chain amino acids, high-density lipoproteins, low-density lipoproteins, glycemic measures, and hypertension. Conclusions: In the large-scale prospective Women's Health Study of 25,315 initially healthy US women followed for 25 years, higher MED intake was associated with approximately one fifth relative risk reduction in mortality. The inverse association was only partially explained by known novel and traditional cardiometabolic factors.

Age of onset and cumulative risk of mental disorders: a cross-national analysis of population surveys from 29 countries.

BACKGROUND: Information on the frequency and timing of mental disorder onsets across the lifespan is of fundamental importance for public health planning. Broad, cross-national estimates of this information from coordinated general population surveys were last updated in 2007. We aimed to provide updated and improved estimates of age-of-onset distributions, lifetime prevalence, and morbid risk. METHODS: In this cross-national analysis, we analysed data from respondents aged 18 years or older to the World Mental Health surveys, a coordinated series of cross-sectional, face-to-face community epidemiological surveys administered between 2001 and 2022. In the surveys, the WHO Composite International Diagnostic Interview, a fully structured psychiatric diagnostic interview, was used to assess age of onset, lifetime prevalence, and morbid risk of 13 DSM-IV mental disorders until age 75 years across surveys by sex. We did not assess ethnicity. The surveys were geographically clustered and weighted to adjust for selection probability, and standard errors of incidence rates and cumulative incidence curves were calculated using the jackknife repeated replications simulation method, taking weighting and geographical clustering of data into account. FINDINGS: We included 156 331 respondents from 32 surveys in 29 countries, including 12 low-income and middle-income countries and 17 high-income countries, and including 85 308 (54·5%) female respondents and 71 023 (45·4%) male respondents. The lifetime prevalence of any mental disorder was 28·6% (95% CI 27·9-29·2) for male respondents and 29·8% (29·2-30·3) for female respondents. Morbid risk of any mental disorder by age 75 years was 46·4% (44·9-47·8) for male respondents and 53·1% (51·9-54·3) for female respondents. Conditional probabilities of first onset peaked at approximately age 15 years, with a median age of onset of 19 years (IQR 14-32) for male respondents and 20 years (12-36) for female respondents. The two most prevalent disorders were alcohol use disorder and major depressive disorder for male respondents and major depressive disorder and specific phobia for female respondents. INTERPRETATION: By age 75 years, approximately half the population can expect to develop one or more of the 13 mental disorders considered in this Article. These disorders typically first emerge in childhood, adolescence, or young adulthood. Services should have the capacity to detect and treat common mental disorders promptly and to optimise care that suits people at these crucial parts of the life course. FUNDING: None.

Evaluation of a Model to Target High-risk Psychiatric Inpatients for an Intensive Postdischarge Suicide Prevention Intervention.

Importance: The months after psychiatric hospital discharge are a time of high risk for suicide. Intensive postdischarge case management, although potentially effective in suicide prevention, is likely to be cost-effective only if targeted at high-risk patients. A previously developed machine learning (ML) model showed that postdischarge suicides can be predicted from electronic health records and geospatial data, but it is unknown if prediction could be improved by adding additional information. Objective: To determine whether model prediction could be improved by adding information extracted from clinical notes and public records. Design, Setting, and Participants: Models were trained to predict suicides in the 12 months after Veterans Health Administration (VHA) short-term (less than 365 days) psychiatric hospitalizations between the beginning of 2010 and September 1, 2012 (299 050 hospitalizations, with 916 hospitalizations followed within 12 months by suicides) and tested in the hospitalizations from September 2, 2012, to December 31, 2013 (149 738 hospitalizations, with 393 hospitalizations followed within 12 months by suicides). Validation focused on net benefit across a range of plausible decision thresholds. Predictor importance was assessed with Shapley additive explanations (SHAP) values. Data were analyzed from January to August 2022. Main Outcomes and Measures: Suicides were defined by the National Death Index. Base model predictors included VHA electronic health records and patient residential data. The expanded predictors came from natural language processing (NLP) of clinical notes and a social determinants of health (SDOH) public records database. Results: The model included 448 788 unique hospitalizations. Net benefit over risk horizons between 3 and 12 months was generally highest for the model that included both NLP and SDOH predictors (area under the receiver operating characteristic curve range, 0.747-0.780; area under the precision recall curve relative to the suicide rate range, 3.87-5.75). NLP and SDOH predictors also had the highest predictor class-level SHAP values (proportional SHAP = 64.0% and 49.3%, respectively), although the single highest positive variable-level SHAP value was for a count of medications classified by the US Food and Drug Administration as increasing suicide risk prescribed the year before hospitalization (proportional SHAP = 15.0%). Conclusions and Relevance: In this study, clinical notes and public records were found to improve ML model prediction of suicide after psychiatric hospitalization. The model had positive net benefit over 3-month to 12-month risk horizons for plausible decision thresholds. Although caution is needed in inferring causality based on predictor importance, several key predictors have potential intervention implications that should be investigated in future studies.

Misuse of DeLong test to compare AUCs for nested models.

The area under the receiver operating characteristics curve (AUC of ROC) is a widely used measure of discrimination in risk prediction models. Routinely, the Mann-Whitney statistics is used as an estimator of AUC, while the change in AUC is tested by the DeLong test. However, very often, in settings where the model is developed and tested on the same dataset, the added predictor is statistically significantly associated with the outcome but fails to produce a significant improvement in the AUC. No conclusive resolution exists to explain this finding. In this paper, we will show that the reason lies in the inappropriate application of the DeLong test in the setting of nested models. Using numerical simulations and a theoretical argument based on generalized U-statistics, we show that if the added predictor is not statistically significantly associated with the outcome, the null distribution is non-normal, contrary to the assumption of DeLong test. Our simulations of different scenarios show that the loss of power because of such a misuse of the DeLong test leads to a conservative test for small and moderate effect sizes. This problem does not exist in cases of predictors that are associated with the outcome and for non-nested models. We suggest that for nested models, only the test of association be performed for the new predictors, and if the result is significant, change in AUC be estimated with an appropriate confidence interval, which can be based on the DeLong approach.

Novel metrics for evaluating improvement in discrimination: net reclassification and integrated discrimination improvement for normal variables and nested models.

Net reclassification and integrated discrimination improvements have been proposed as alternatives to the increase in the area under the curve for evaluating improvement in the performance of risk assessment algorithms introduced by the addition of new phenotypic or genetic markers. In this paper, we demonstrate that in the setting of linear discriminant analysis, under the assumptions of multivariate normality, all three measures can be presented as functions of the squared Mahalanobis distance. This relationship affords an interpretation of the magnitude of these measures in the familiar language of effect size for uncorrelated variables. Furthermore, it allows us to conclude that net reclassification improvement can be viewed as a universal measure of effect size. Our theoretical developments are illustrated with an example based on the Framingham Heart Study risk assessment model for high-risk men in primary prevention of cardiovascular disease.

Quantitative Comparison of Statistical Methods for Analyzing Human Metabolomics Data.

Emerging technologies now allow for mass spectrometry-based profiling of thousands of small molecule metabolites ('metabolomics') in an increasing number of biosamples. While offering great promise for insight into the pathogenesis of human disease, standard approaches have not yet been established for statistically analyzing increasingly complex, high-dimensional human metabolomics data in relation to clinical phenotypes, including disease outcomes. To determine optimal approaches for analysis, we formally compare traditional and newer statistical learning methods across a range of metabolomics dataset types. In simulated and experimental metabolomics data derived from large population-based human cohorts, we observe that with an increasing number of study subjects, univariate compared to multivariate methods result in an apparently higher false discovery rate as represented by substantial correlation between metabolites directly associated with the outcome and metabolites not associated with the outcome. Although the higher frequency of such associations would not be considered false in the strict statistical sense, it may be considered biologically less informative. In scenarios wherein the number of assayed metabolites increases, as in measures of nontargeted versus targeted metabolomics, multivariate methods performed especially favorably across a range of statistical operating characteristics. In nontargeted metabolomics datasets that included thousands of metabolite measures, sparse multivariate models demonstrated greater selectivity and lower potential for spurious relationships. When the number of metabolites was similar to or exceeded the number of study subjects, as is common with nontargeted metabolomics analysis of relatively small cohorts, sparse multivariate models exhibited the most-robust statistical power with more consistent results. These findings have important implications for metabolomics analysis in human disease.

Equivalence of improvement in area under ROC curve and linear discriminant analysis coefficient under assumption of normality.

In this paper we investigate the addition of new variables to an existing risk prediction model and the subsequent impact on discrimination quantified by the area under the receiver operating characteristics curve (AUC of ROC). Based on practical experience, concerns have emerged that the significance of association of the variable under study with the outcome in the risk model does not correspond to the significance of the change in AUC: that is, often the variable is significant, but the change in AUC is not. This paper demonstrates that under the assumption of multivariate normality and employing linear discriminant analysis (LDA) to construct the risk prediction tool, statistical significance of the new predictor(s) is equivalent to the statistical significance of the increase in AUC. Under these assumptions the result extends asymptotically to logistic regression. We further show that equality of variance-covariance matrices of predictors within cases and non-cases is not necessary when LDA is used. However, our practical example from the Framingham Heart Study data suggests that the finding might be sensitive to the assumption of normality.

Association of Lipid, Inflammatory, and Metabolic Biomarkers With Age at Onset for Incident Coronary Heart Disease in Women.

Importance: Risk profiles for premature coronary heart disease (CHD) are unclear. Objective: To examine baseline risk profiles for incident CHD in women by age at onset. Design, Setting, and Participants: A prospective cohort of US female health professionals participating in the Women's Health Study was conducted; median follow-up was 21.4 years. Participants included 28 024 women aged 45 years or older without known cardiovascular disease. Baseline profiles were obtained from April 30, 1993, to January 24, 1996, and analyses were conducted from October 1, 2017, to October 1, 2020. Exposures: More than 50 clinical, lipid, inflammatory, and metabolic risk factors and biomarkers. Main Outcomes and Measures: Four age groups were examined (<55, 55 to <65, 65 to <75, and ≥75 years) for CHD onset, and adjusted hazard ratios (aHRs) were calculated using stratified Cox proportional hazard regression models with age as the time scale and adjusting for clinical factors. Women contributed to different age groups over time. Results: Of the clinical factors in the women, diabetes had the highest aHR for CHD onset at any age, ranging from 10.71 (95% CI, 5.57-20.60) at CHD onset in those younger than 55 years to 3.47 (95% CI, 2.47-4.87) at CHD onset in those 75 years or older. Risks that were also noted for CHD onset in participants younger than 55 years included metabolic syndrome (aHR, 6.09; 95% CI, 3.60-10.29), hypertension (aHR, 4.58; 95% CI, 2.76-7.60), obesity (aHR, 4.33; 95% CI, 2.31-8.11), and smoking (aHR, 3.92; 95% CI, 2.32-6.63). Myocardial infarction in a parent before age 60 years was associated with 1.5- to 2-fold risk of CHD in participants up to age 75 years. From approximately 50 biomarkers, lipoprotein insulin resistance had the highest standardized aHR: 6.40 (95% CI, 3.14-13.06) for CHD onset in women younger than 55 years, attenuating with age. In comparison, weaker but significant associations with CHD in women younger than 55 years were noted (per SD increment) for low-density lipoprotein cholesterol (aHR, 1.38; 95% CI, 1.10-1.74), non-high-density lipoprotein cholesterol (aHR, 1.67; 95% CI, 1.36-2.04), apolipoprotein B (aHR, 1.89; 95% CI, 1.52-2.35), triglycerides (aHR, 2.14; 95% CI, 1.72-2.67), and inflammatory biomarkers (1.2- to 1.8-fold)-all attenuating with age. Some biomarkers had similar CHD age associations (eg, physical inactivity, lipoprotein[a], total high-density lipoprotein particles), while a few had no association with CHD onset at any age. Most risk factors and biomarkers had associations that attenuated with increasing age at onset. Conclusions and Relevance: In this cohort study, diabetes and insulin resistance, in addition to hypertension, obesity, and smoking, appeared to be the strongest risk factors for premature onset of CHD. Most risk factors had attenuated relative rates at older ages.