Data Resource Profile: Self-reported data in the Million Veteran Program: survey development and insights from the first 850 736 participants

Protein Identification for Stroke Progression via Mendelian Randomization in Million Veteran Program and UK Biobank

BACKGROUND

Individuals who have experienced a stroke, or transient ischemic attack, face a heightened risk of future cardiovascular events. Identification of genetic and molecular risk factors for subsequent cardiovascular outcomes may identify effective therapeutic targets to improve prognosis after an incident stroke.

METHODS

We performed genome-wide association studies for subsequent major adverse cardiovascular events (MACE; ncases=51 929; ncontrols=39 980) and subsequent arterial ischemic stroke (AIS; ncases=45 120; ncontrols=46 789) after the first incident stroke within the Million Veteran Program and UK Biobank. We then used genetic variants associated with proteins (protein quantitative trait loci) to determine the effect of 1463 plasma protein abundances on subsequent MACE using Mendelian randomization.

RESULTS

Two variants were significantly associated with subsequent cardiovascular events: rs76472767 near gene RNF220 (odds ratio, 0.75 [95% CI, 0.64-0.85]; P=3.69×10-8) with subsequent AIS and rs13294166 near gene LINC01492 (odds ratio, 1.52 [95% CI, 1.37-1.67]; P=3.77×10-8) with subsequent MACE. Using Mendelian randomization, we identified 2 proteins with an effect on subsequent MACE after a stroke: CCL27 ([C-C motif chemokine 27], effect odds ratio, 0.77 [95% CI, 0.66-0.88]; adjusted P=0.05) and TNFRSF14 ([tumor necrosis factor receptor superfamily member 14], effect odds ratio, 1.42 [95% CI, 1.24-1.60]; adjusted P=0.006). These proteins are not associated with incident AIS and are implicated to have a role in inflammation.

CONCLUSIONS

We found evidence that 2 proteins with little effect on incident stroke appear to influence subsequent MACE after incident AIS. These associations suggest that inflammation is a contributing factor to subsequent MACE outcomes after incident AIS and highlights potential novel targets.

Diversity and scale: Genetic architecture of 2068 traits in the VA Million Veteran Program

One of the justifiable criticisms of human genetic studies is the underrepresentation of participants from diverse populations. Lack of inclusion must be addressed at-scale to identify causal disease factors and understand the genetic causes of health disparities. We present genome-wide associations for 2068 traits from 635,969 participants in the Department of Veterans Affairs Million Veteran Program, a longitudinal study of diverse United States Veterans. Systematic analysis revealed 13,672 genomic risk loci; 1608 were only significant after including non-European populations. Fine-mapping identified causal variants at 6318 signals across 613 traits. One-third (n = 2069) were identified in participants from non-European populations. This reveals a broadly similar genetic architecture across populations, highlights genetic insights gained from underrepresented groups, and presents an extensive atlas of genetic associations.

Subjective cognitive concerns, APOE ε4, PTSD symptoms, and risk for dementia among older veterans

BACKGROUND

Posttraumatic stress disorder (PTSD) and traumatic brain injury (TBI) are associated with self-reported problems with cognition as well as risk for Alzheimer's disease and related dementias (ADRD). Overlapping symptom profiles observed in cognitive disorders, psychiatric disorders, and environmental exposures (e.g., head injury) can complicate the detection of early signs of ADRD. The interplay between PTSD, head injury, subjective (self-reported) cognitive concerns and genetic risk for ADRD is also not well understood, particularly in diverse ancestry groups.

METHODS

Using data from the U.S. Department of Veterans Affairs (VA) Million Veteran Program (MVP), we examined the relationship between dementia risk factors (APOE ε4, PTSD, TBI) and subjective cognitive concerns (SCC) measured in individuals of European (n = 140,921), African (n = 15,788), and Hispanic (n = 8,064) ancestry (EA, AA, and HA, respectively). We then used data from the VA electronic medical record to perform a retrospective survival analysis evaluating PTSD, TBI, APOE ε4, and SCC and their associations with risk of conversion to ADRD in Veterans aged 65 and older.

RESULTS

PTSD symptoms (B = 0.50-0.52, p < 1E-250) and probable TBI (B = 0.05-0.19, p = 1.51E-07 - 0.002) were positively associated with SCC across all three ancestry groups. APOE ε4 was associated with greater SCC in EA Veterans aged 65 and older (B = 0.037, p = 1.88E-12). Results of Cox models indicated that PTSD symptoms (hazard ratio [HR] = 1.13-1.21), APOE ε4 (HR = 1.73-2.05) and SCC (HR = 1.18-1.37) were positively associated with risk for ADRD across all three ancestry groups. In the EA group, probable TBI also contributed to increased risk of ADRD (HR = 1.18).

CONCLUSIONS

The findings underscore the value of SCC as an indicator of ADRD risk in Veterans 65 and older when considered in conjunction with other influential genetic, clinical, and demographic risk factors.

Accelerating Genome- and Phenome-Wide Association Studies using GPUs - A case study using data from the Million Veteran Program

The expansion of biobanks has significantly propelled genomic discoveries yet the sheer scale of data within these repositories poses formidable computational hurdles, particularly in handling extensive matrix operations required by prevailing statistical frameworks. In this work, we introduce computational optimizations to the SAIGE (Scalable and Accurate Implementation of Generalized Mixed Model) algorithm, notably employing a GPU-based distributed computing approach to tackle these challenges. We applied these optimizations to conduct a large-scale genome-wide association study (GWAS) across 2,068 phenotypes derived from electronic health records of 635,969 diverse participants from the Veterans Affairs (VA) Million Veteran Program (MVP). Our strategies enabled scaling up the analysis to over 6,000 nodes on the Department of Energy (DOE) Oak Ridge Leadership Computing Facility (OLCF) Summit High-Performance Computer (HPC), resulting in a 20-fold acceleration compared to the baseline model. We also provide a Docker container with our optimizations that was successfully used on multiple cloud infrastructures on UK Biobank and All of Us datasets where we showed significant time and cost benefits over the baseline SAIGE model.

Protein identification for stroke progression via Mendelian Randomization in Million Veteran Program and UK Biobank

Background

Individuals who have experienced a stroke, or transient ischemic attack, face a heightened risk of future cardiovascular events. Identification of genetic and molecular risk factors for subsequent cardiovascular outcomes may identify effective therapeutic targets to improve prognosis after an incident stroke.

Methods

We performed genome-wide association studies (GWAS) for subsequent major adverse cardiovascular events (MACE) (Ncases=51,929, Ncntrl=39,980) and subsequent arterial ischemic stroke (AIS) Ncases=45,120, Ncntrl=46,789) after first incident stroke within the Million Veteran Program and UK Biobank. We then used genetic variants associated with proteins (pQTLs) to determine the effect of 1,463 plasma protein abundances on subsequent MACE using Mendelian randomization (MR).

Results

Two variants were significantly associated with subsequent cardiovascular events: rs76472767 (OR=0.75, 95% CI = 0.64-0.85, p= 3.69×10-08) with subsequent AIS and rs13294166 (OR=1.52, 95% CI = 1.37-1.67, p=3.77×10-08) with subsequent MACE. Using MR, we identified 2 proteins with an effect on subsequent MACE after a stroke: CCL27 (effect OR= 0.77, 95% CI = 0.66-0.88, adj. p=0.05), and TNFRSF14 (effect OR=1.42, 95% CI = 1.24-1.60, adj. p=0.006). These proteins are not associated with incident AIS and are implicated to have a role in inflammation.

Conclusions

We found evidence that two proteins with little effect on incident stroke appear to influence subsequent MACE after incident AIS. These associations suggest that inflammation is a contributing factor to subsequent MACE outcomes after incident AIS and highlights potential novel targets.

The Million Veteran Program 1990-1991 Gulf War Era Survey: An Evaluation of Veteran Response, Characteristics, and Representativeness of the Gulf War Era Veteran Population

To address gaps in understanding the pathophysiology of Gulf War Illness (GWI), the VA Million Veteran Program (MVP) developed and implemented a survey to MVP enrollees who served in the U.S. military during the 1990-1991 Persian Gulf War (GW). Eligible Veterans were invited via mail to complete a survey assessing health conditions as well as GW-specific deployment characteristics and exposures. We evaluated the representativeness of this GW-era cohort relative to the broader population by comparing demographic, military, and health characteristics between respondents and non-respondents, as well as with all GW-era Veterans who have used Veterans Health Administration (VHA) services and the full population of U.S. GW-deployed Veterans. A total of 109,976 MVP GW-era Veterans were invited to participate and 45,270 (41%) returned a completed survey. Respondents were 84% male, 72% White, 8% Hispanic, with a mean age of 61.6 years (SD = 8.5). Respondents were more likely to be older, White, married, better educated, slightly healthier, and have higher socioeconomic status than non-respondents, but reported similar medical conditions and comparable health status. Although generally similar to all GW-era Veterans using VHA services and the full population of U.S. GW Veterans, respondents included higher proportions of women and military officers, and were slightly older. In conclusion, sample characteristics of the MVP GW-era cohort can be considered generally representative of the broader GW-era Veteran population. The sample represents the largest research cohort of GW-era Veterans established to date and provides a uniquely valuable resource for conducting in-depth studies to evaluate health conditions affecting 1990-1991 GW-era Veterans.

Impact of 8 lifestyle factors on mortality and life expectancy among United States veterans: The Million Veteran Program

BACKGROUND

Lifestyle medicine has been proposed as a way to address the root causes of chronic disease and their associated health care costs.

OBJECTIVE

This study aimed to estimate mortality risk and longevity associated with individual lifestyle factors and comprehensive lifestyle therapy.

METHODS

Age- and sex-specific mortality rates were calculated on the basis of 719,147 veterans aged 40-99 y enrolled in the Veteran Affairs Million Veteran Program (2011-2019). Hazard ratios and estimated increase in life expectancy were examined among a subgroup of 276,132 veterans with complete data on 8 lifestyle factors at baseline. The 8 lifestyle factors included never smoking, physical activity, no excessive alcohol consumption, restorative sleep, nutrition, stress management, social connections, and no opioid use disorder.

RESULTS

On the basis of 1.12 million person-years of follow-up, 34,247 deaths were recorded. Among veterans who adopted 1, 2, 3, 4, 5, 6, 7, and 8 lifestyle factors, the adjusted hazard ratios for mortality were 0.74 (0.60-0.90), 0.60 (95% CI: 0.49, 0.73), 0.50 (95% CI: 0.41, 0.61), 0.43 (95% CI: 0.35, 0.52), 0.35 (95% CI: 0.29, 0.43), 0.27 (95% CI: 0.22, 0.33), 0.21 (95% CI: 0.17, 0.26), and 0.13 (95% CI: 0.10, 0.16), respectively, as compared with veterans with no adopted lifestyle factors. The estimated life expectancy at age 40 y was 23.0, 26.5, 28.8, 30.8, 32.7, 35.1, 38.3, 41.3, and 47.0 y among males and 27.0, 28.8, 33.1, 38.0, 39.2, 41.4, 43.8, 46.3, and 47.5 y for females who adopted 0, 1, 2, 3, 4, 5, 6, 7, and 8 lifestyle factors, respectively. The difference in life expectancy at age 40 y was 24.0 y for male veterans and 20.5 y for female veterans when comparing adoption of 8-9 lifestyle factors.

CONCLUSIONS

A combination of 8 lifestyle factors is associated with a significantly lower risk of premature mortality and an estimated prolonged life expectancy.

Whole Person Modeling: a transdisciplinary approach to mental health research

The growing global burden of mental illness has prompted calls for innovative research strategies. Theoretical models of mental health include complex contributions of biological, psychosocial, experiential, and other environmental influences. Accordingly, neuropsychiatric research has self-organized into largely isolated disciplines working to decode each individual contribution. However, research directly modeling objective biological measurements in combination with cognitive, psychological, demographic, or other environmental measurements is only now beginning to proliferate. This review aims to (1) to describe the landscape of modern mental health research and current movement towards integrative study, (2) to provide a concrete framework for quantitative integrative research, which we call Whole Person Modeling, (3) to explore existing and emerging techniques and methods used in Whole Person Modeling, and (4) to discuss our observations about the scarcity, potential value, and untested aspects of highly transdisciplinary research in general. Whole Person Modeling studies have the potential to provide a better understanding of multilevel phenomena, deliver more accurate diagnostic and prognostic tests to aid in clinical decision making, and test long standing theoretical models of mental illness. Some current barriers to progress include challenges with interdisciplinary communication and collaboration, systemic cultural barriers to transdisciplinary career paths, technical challenges in model specification, bias, and data harmonization, and gaps in transdisciplinary educational programs. We hope to ease anxiety in the field surrounding the often mysterious and intimidating world of transdisciplinary, data-driven mental health research and provide a useful orientation for students or highly specialized researchers who are new to this area.

Diversity and Scale: Genetic Architecture of 2,068 Traits in the VA Million Veteran Program

Genome-wide association studies (GWAS) have underrepresented individuals from non-European populations, impeding progress in characterizing the genetic architecture and consequences of health and disease traits. To address this, we present a population-stratified phenome-wide GWAS followed by a multi-population meta-analysis for 2,068 traits derived from electronic health records of 635,969 participants in the Million Veteran Program (MVP), a longitudinal cohort study of diverse U.S. Veterans genetically similar to the respective African (121,177), Admixed American (59,048), East Asian (6,702), and European (449,042) superpopulations defined by the 1000 Genomes Project. We identified 38,270 independent variants associating with one or more traits at experiment-wide P < 4.6 × 10 - 11 significance; fine-mapping 6,318 signals identified from 613 traits to single-variant resolution. Among these, a third (2,069) of the associations were found only among participants genetically similar to non-European reference populations, demonstrating the importance of expanding diversity in genetic studies. Our work provides a comprehensive atlas of phenome-wide genetic associations for future studies dissecting the architecture of complex traits in diverse populations.

Genome-Wide Investigation of Maximum Habitual Alcohol Intake in US Veterans in Relation to Alcohol Consumption Traits and Alcohol Use Disorder

Importance

Alcohol genome-wide association studies (GWASs) have generally focused on alcohol consumption and alcohol use disorder (AUD); few have examined habitual drinking behaviors like maximum habitual alcohol intake (MaxAlc).

Objectives

To identify genetic loci associated with MaxAlc and to elucidate the genetic architecture across alcohol traits.

Design, Setting, and Participants

This MaxAlc genetic association study was performed among Million Veteran Program participants enrolled from January 10, 2011, to September 30, 2020. Ancestry-specific GWASs were conducted in participants with European (n = 218 623) and African (n = 29 132) ancestry, then meta-analyzed (N = 247 755). Linkage-disequilibrium score regression was used to estimate single nucleotide variant (SNV)-heritability and genetic correlations (rg) with other alcohol and psychiatric traits. Genomic structural equation modeling (gSEM) was used to evaluate genetic associations between MaxAlc and other alcohol traits. Mendelian randomization was used to examine potential causal relationships between MaxAlc and liver enzyme levels. MTAG (multitrait analysis of GWAS) was used to analyze MaxAlc and problematic alcohol use (PAU) jointly.

Exposures

Genetic associations.

Main Outcomes and Measures

MaxAlc was defined from the following survey item: "in a typical month, what is/was the largest number of drinks of alcohol you may have had in one day?" with ordinal responses from 0 to 15 or more drinks.

Results

GWASs were conducted on sample sizes of as many as 247 455 US veterans. Participants were 92.68% male and had mean (SD) age of 65.92 (11.70) years. The MaxAlc GWAS resulted in 15 genome-wide significant loci. Top associations in European-ancestry and African-ancestry participants were with known functional variants in the ADH1B gene, namely rs1229984 (P = 3.12 × 10-101) and rs2066702 (P = 6.30 × 10-17), respectively. Novel associations were also found. SNV-heritability was 6.65% (SE, 0.41) in European-ancestry participants and 3.42% (SE, 1.46) in African-ancestry participants. MaxAlc was positively correlated with PAU (rg = 0.79; P = 3.95 × 10-149) and AUD (rg = 0.76; P = 1.26 × 10-127) and had negative rg with the UK Biobank "alcohol usually taken with meals" (rg = -0.53; P = 1.40 × 10-50). For psychiatric traits, MaxAlc had the strongest genetic correlation with suicide attempt (rg = 0.40; P = 3.02 × 10-21). gSEM supported a 2-factor model with MaxAlc loading on a factor with PAU and AUD and other alcohol consumption measures loading on a separate factor. Mendelian randomization supported an association between MaxAlc and the liver enzyme gamma-glutamyltransferase (β = 0.012; P = 2.66 × 10-10). MaxAlc MTAG resulted in 31 genome-wide significant loci.

Conclusions and Relevance

The findings suggest that MaxAlc closely aligns genetically with PAU traits. This study improves understanding of the mechanisms associated with normative alcohol consumption vs problematic habitual use and AUD as well as how MaxAlc relates to psychiatric and medical conditions genetically and biologically.