High dimensional predictions of suicide risk in 4.2 million US Veterans using ensemble transfer learning.

We present an ensemble transfer learning method to predict suicide from Veterans Affairs (VA) electronic medical records (EMR). A diverse set of base models was trained to predict a binary outcome constructed from reported suicide, suicide attempt, and overdose diagnoses with varying choices of study design and prediction methodology. Each model used twenty cross-sectional and 190 longitudinal variables observed in eight time intervals covering 7.5 years prior to the time of prediction. Ensembles of seven base models were created and fine-tuned with ten variables expected to change with study design and outcome definition in order to predict suicide and combined outcome in a prospective cohort. The ensemble models achieved c-statistics of 0.73 on 2-year suicide risk and 0.83 on the combined outcome when predicting on a prospective cohort of [Formula: see text] 4.2 M veterans. The ensembles rely on nonlinear base models trained using a matched retrospective nested case-control (Rcc) study cohort and show good calibration across a diversity of subgroups, including risk strata, age, sex, race, and level of healthcare utilization. In addition, a linear Rcc base model provided a rich set of biological predictors, including indicators of suicide, substance use disorder, mental health diagnoses and treatments, hypoxia and vascular damage, and demographics.

GWAS Meta-Analysis of Suicide Attempt: Identification of 12 Genome-Wide Significant Loci and Implication of Genetic Risks for Specific Health Factors.

OBJECTIVE: Suicidal behavior is heritable and is a major cause of death worldwide. Two large-scale genome-wide association studies (GWASs) recently discovered and cross-validated genome-wide significant (GWS) loci for suicide attempt (SA). The present study leveraged the genetic cohorts from both studies to conduct the largest GWAS meta-analysis of SA to date. Multi-ancestry and admixture-specific meta-analyses were conducted within groups of significant African, East Asian, and European ancestry admixtures. METHODS: This study comprised 22 cohorts, including 43,871 SA cases and 915,025 ancestry-matched controls. Analytical methods across multi-ancestry and individual ancestry admixtures included inverse variance-weighted fixed-effects meta-analyses, followed by gene, gene-set, tissue-set, and drug-target enrichment, as well as summary-data-based Mendelian randomization with brain expression quantitative trait loci data, phenome-wide genetic correlation, and genetic causal proportion analyses. RESULTS: Multi-ancestry and European ancestry admixture GWAS meta-analyses identified 12 risk loci at p values <5×10-8. These loci were mostly intergenic and implicated DRD2, SLC6A9, FURIN, NLGN1, SOX5, PDE4B, and CACNG2. The multi-ancestry SNP-based heritability estimate of SA was 5.7% on the liability scale (SE=0.003, p=5.7×10-80). Significant brain tissue gene expression and drug set enrichment were observed. There was shared genetic variation of SA with attention deficit hyperactivity disorder, smoking, and risk tolerance after conditioning SA on both major depressive disorder and posttraumatic stress disorder. Genetic causal proportion analyses implicated shared genetic risk for specific health factors. CONCLUSIONS: This multi-ancestry analysis of suicide attempt identified several loci contributing to risk and establishes significant shared genetic covariation with clinical phenotypes. These findings provide insight into genetic factors associated with suicide attempt across ancestry admixture populations, in veteran and civilian populations, and in attempt versus death.

Identification of Novel, Replicable Genetic Risk Loci for Suicidal Thoughts and Behaviors Among US Military Veterans.

Importance: Suicide is a leading cause of death; however, the molecular genetic basis of suicidal thoughts and behaviors (SITB) remains unknown. Objective: To identify novel, replicable genomic risk loci for SITB. Design, Setting, and Participants: This genome-wide association study included 633 778 US military veterans with and without SITB, as identified through electronic health records. GWAS was performed separately by ancestry, controlling for sex, age, and genetic substructure. Cross-ancestry risk loci were identified through meta-analysis. Study enrollment began in 2011 and is ongoing. Data were analyzed from November 2021 to August 2022. Main Outcome and Measures: SITB. Results: A total of 633 778 US military veterans were included in the analysis (57 152 [9%] female; 121 118 [19.1%] African ancestry, 8285 [1.3%] Asian ancestry, 452 767 [71.4%] European ancestry, and 51 608 [8.1%] Hispanic ancestry), including 121 211 individuals with SITB (19.1%). Meta-analysis identified more than 200 GWS (P < 5 × 10-8) cross-ancestry risk single-nucleotide variants for SITB concentrated in 7 regions on chromosomes 2, 6, 9, 11, 14, 16, and 18. Top single-nucleotide variants were largely intronic in nature; 5 were independently replicated in ISGC, including rs6557168 in ESR1, rs12808482 in DRD2, rs77641763 in EXD3, rs10671545 in DCC, and rs36006172 in TRAF3. Associations for FBXL19 and AC018880.2 were not replicated. Gene-based analyses implicated 24 additional GWS cross-ancestry risk genes, including FURIN, TSNARE1, and the NCAM1-TTC12-ANKK1-DRD2 gene cluster. Cross-ancestry enrichment analyses revealed significant enrichment for expression in brain and pituitary tissue, synapse and ubiquitination processes, amphetamine addiction, parathyroid hormone synthesis, axon guidance, and dopaminergic pathways. Seven other unique European ancestry-specific GWS loci were identified, 2 of which (POM121L2 and METTL15/LINC02758) were replicated. Two additional GWS ancestry-specific loci were identified within the African ancestry (PET112/GATB) and Hispanic ancestry (intergenic locus on chromosome 4) subsets, both of which were replicated. No GWS loci were identified within the Asian ancestry subset; however, significant enrichment was observed for axon guidance, cyclic adenosine monophosphate signaling, focal adhesion, glutamatergic synapse, and oxytocin signaling pathways across all ancestries. Within the European ancestry subset, genetic correlations (r > 0.75) were observed between the SITB phenotype and a suicide attempt-only phenotype, depression, and posttraumatic stress disorder. Additionally, polygenic risk score analyses revealed that the Million Veteran Program polygenic risk score had nominally significant main effects in 2 independent samples of veterans of European and African ancestry. Conclusions and Relevance: The findings of this analysis may advance understanding of the molecular genetic basis of SITB and provide evidence for ESR1, DRD2, TRAF3, and DCC as cross-ancestry candidate risk genes. More work is needed to replicate these findings and to determine if and how these genes might impact clinical care.

Perspectives and Practice in the Identification and Treatment of Opioid Use, Alcohol Use, and Depressive Disorders.

OBJECTIVE: Little research has focused on the treatment of adults with substance use disorders in primary care despite the high occurrence, morbidity, and mortality associated with these disorders. METHODS: An electronic survey was administered to primary care providers in a large health system to assess screening and treatment practices and comfort managing opioid use, alcohol use, and depressive disorders. A total of 146 providers completed the survey (32%). RESULTS: Providers were significantly less likely to screen for or treat opioid use disorders and alcohol use disorders, compared with depression. Providers reported feeling significantly less confident, less prepared, less expected to treat, less sure of the appropriateness of treating, and less able to navigate community resources in the treatment of opioid and alcohol use disorders, compared with depression. CONCLUSIONS: Given the preponderance of substance use disorders in primary care, increased attention to equipping primary care providers to treat these conditions is warranted.

Projected Prevalence of Actionable Pharmacogenetic Variants and Level A Drugs Prescribed Among US Veterans Health Administration Pharmacy Users.

Importance: Implementation of pharmacogenetic testing to guide drug prescribing has potential to improve drug response and prevent adverse events. Robust data exist for more than 30 gene-drug pairs linking genotype to drug response phenotypes; however, it is unclear which pharmacogenetic tests, if implemented, would provide the greatest utility for a given patient population. Objectives: To project the proportion of veterans in the US Veterans Health Administration (VHA) with actionable pharmacogenetic variants and evaluate how testing might be associated with prescribing decisions. Design, Setting, and Participants: This cross-sectional study included veterans who used national VHA pharmacy services from October 1, 2011, to September 30, 2017. Data analyses began April 26, 2018, and were completed February 6, 2019. Exposures: Receipt of level A drugs based on VHA pharmacy dispensing records. Main Outcomes and Measures: Projected prevalence of actionable pharmacogenetic variants among VHA pharmacy users based on variant frequencies from the 1000 Genomes Project and veteran demographic characteristics; incident number of level A prescriptions, and proportion of new level A drug recipients projected to carry an actionable pharmacogenetic variant. Results: During the study, 7 769 359 veterans (mean [SD] age, 58.1 [17.8] years; 7 021 504 [90.4%] men) used VHA pharmacy services. It was projected that 99% of VHA pharmacy users would carry at least 1 actionable pharmacogenetic variant. Among VHA pharmacy users, 4 259 153 (54.8%) received at least 1 level A drug with 1 188 124 (15.3%) receiving 2 drugs, and 912 189 (11.7%) receiving 3 or more drugs. The most common incident prescriptions during the study were tramadol (923 671 new recipients), simvastatin (533 928 new recipients), citalopram (266 952 new recipients), and warfarin (205 177 new recipients). Gene-drug interactions projected to have substantial clinical impacts in the VHA population include the interaction of SLCO1B1 with simvastatin (1 988 956 veterans [25.6%]), CYP2D6 with tramadol (318 544 veterans [4.1%]), and CYP2C9 or VKORC1 with warfarin (7 163 349 veterans [92.2%]). Conclusions and Relevance: Clinically important pharmacogenetic variants are highly prevalent in the VHA population. Almost all veterans would carry an actionable variant, and more than half of the population had been exposed to a drug affected by these variants. These results suggest that pharmacogenetic testing has the potential to affect pharmacotherapy decisions for commonly prescribed outpatient medications for many veterans.

Mental health symptom severity in cannabis using and non-using Veterans with probable PTSD.

BACKGROUND: Posttraumatic Stress Disorder (PTSD) is a disabling illness suffered by many Veterans returning from war. Some Veterans believe that cannabis may be therapeutic for PTSD. The purpose of this study was to better understand the association between cannabis use and PTSD symptoms. METHODS: The study was a matched case-control cross-sectional evaluation of the psychiatric and sociocultural associations of cannabis use in Veterans with probable PTSD. Patient self-report measures were examined comparing cannabis users (cases) to non-users (controls) who were case-matched on age and gender. RESULTS: Results indicated that there were no significant differences between cases and controls in mean PTSD Checklist-Civilian version (PCL-C) scores (59.2 and 59.1, respectively). There was also no association between PTSD scores and frequency of cannabis use. It was also observed that cases were more likely to be non-Caucasian, financially challenged, and unmarried. LIMITATIONS: The sample is a convenience sample of Veterans being referred for a clinical assessment and therefore, sampling biases may limit the generalizability of the results to other populations including Veterans not seeking health care in the Veterans Affairs (VA) system. CONCLUSIONS: The results do not support the theory that cannabis use would be associated with less severe PTSD symptoms. Results do suggest important sociocultural differences in cannabis users compared to controls.

Suicidal ideation in veterans misusing alcohol: relationships with insomnia symptoms and sleep duration.

OBJECTIVE: The aim of this investigation was to assess the relationships between suicidal ideation and insomnia symptoms in Veterans misusing alcohol. METHOD: Data were extracted in this retrospective chart review of Veterans referred from primary care for a behavioral health evaluation (N=161) based on evidence of heavy drinking, drug use or another behavioral problem. Suicidal ideation (SI) was assessed using the Paykel questionnaire. Insomnia symptoms were assessed with standard diary questions in an interview format and pertained to sleep latency (SL), wake after sleep onset time (WASO), sleep quality (SQ), and habitual sleep duration (HSD). The relations between suicidal ideation and insomnia symptoms were assessed using ordinal regression analyses adjusted for socio-demographic, psychiatric and addiction-related variables. RESULTS: Suicidal ideation was reported in 62 (39%) of the Veterans interviewed. In a multivariable model, only inadequate SQ was associated with suicidal ideation. Short sleepers were more likely to endorse suicidal ideation and have attempted suicide in the past year. In addition, older age, inadequate financial status, and the presence of a psychiatric disorder were also significantly associated with suicidal ideation in most of the adjusted models. CONCLUSION: Given their association with suicidal ideation, insomnia symptoms in Veterans misusing alcohol should prompt an assessment of underlying psychiatric and social factors.

Minimal Erythema Dose (MED) testing.

Ultraviolet radiation (UV) therapy is sometimes used as a treatment for various common skin conditions, including psoriasis, acne, and eczema. The dosage of UV light is prescribed according to an individual's skin sensitivity. Thus, to establish the proper dosage of UV light to administer to a patient, the patient is sometimes screened to determine a minimal erythema dose (MED), which is the amount of UV radiation that will produce minimal erythema (sunburn or redness caused by engorgement of capillaries) of an individual's skin within a few hours following exposure. This article describes how to conduct minimal erythema dose (MED) testing. There is currently no easy way to determine an appropriate UV dose for clinical or research purposes without conducting formal MED testing, requiring observation hours after testing, or informal trial and error testing with the risks of under- or over-dosing. However, some alternative methods are discussed.

Insomnia in alcohol dependence: predictors of symptoms in a sample of veterans referred from primary care.

OBJECTIVE: Patients with alcohol dependence presenting for treatment may have multiple associated co-morbid conditions and limited social supports, which complicate treatment. Each of these factors has been independently associated with complaints of insomnia. In this preliminary study, we investigated the relations between insomnia complaints and socio-demographic factors and psychiatric co-morbidity in treatment-seeking patients with alcohol dependence. METHOD: We conducted a retrospective chart review on 84 consecutive patients referred to the Behavioral Health Laboratory of the Philadelphia Veterans Affairs Medical Center for evaluation of psychiatric and substance use disorders. Patients met DSM-IV diagnostic criteria for alcohol dependence and completed a series of self-assessments of sleep. Univariate and multivariable analyses were used to examine the relations amongst the variables of interest. RESULTS: In multivariable models, Sleep Latency was significantly greater in individuals without partners (p = .01), those with psychiatric disorders (p = .03) and smokers (p = .01), with a non-significant trend for those with past-year suicidal ideation. No significant predictor of Wake Time After Sleep Onset was seen. Poor Sleep Quality was predicted by younger age (OR = .93 [.88, .98], p = .004) and the presence of a psychiatric disorder (OR = 20.80 [4, 102], p = .0002), with a non-significant trend for suicidal ideation. CONCLUSIONS: Insomnia symptoms in treatment-seeking alcohol dependent patients should prompt consideration of the individuals' psychiatric and psychosocial features.

Low frequency genetic variants in the µ-opioid receptor (OPRM1) affect risk for addiction to heroin and cocaine.

The µ-opioid receptor (MOR) binds exogenous and endogenous opioids and is known to mediate the rewarding effects of drugs of abuse. Numerous genetic studies have sought to identify common genetic variation in the gene encoding MOR (OPRM1) that affects risk for drug addiction. The purpose of this study was to examine the contribution of rare coding variants in OPRM1 to the risk for addiction. Rare and low frequency variants were selected using the National Heart Lung and Blood Institute - Exome Sequencing Project (NHLBI-ESP) database, which has screened the exomes of over 6500 individuals. Two SNPs (rs62638690 and rs17174794) were selected for genotyping in 1377 European American individuals addicted to heroin and/or cocaine. Two different SNPs (rs1799971 and rs17174801) were genotyped in 1238 African American individuals addicted to heroin and/or cocaine. Using the minor allele frequencies from the NHLBI-ESP dataset as a comparison group, case-control association analyses were performed. Results revealed an association between rs62638690 and cocaine and heroin addiction in European Americans (p=0.02; 95% C.I. 0.47 [0.24-0.92]). This study suggests a potential role for rare OPRM1 variants in addiction disorders and highlights an area worthy of future study.

Identifying risk factors for marijuana use among veterans affairs patients.

OBJECTIVES: Cannabis is the most widely used drug in the United States, and its use carries negative health consequences; however, universal screening for cannabis use is cumbersome. If data commonly collected in the primary care setting (eg, use of alcohol, smoking status, and depression symptoms) could predict cannabis use, then providers can implement targeted marijuana screening in high-risk groups. METHODS: We reviewed Behavioral Health Laboratory data collected between 2003 and 2006 from 5512 patients referred by Veterans Affairs primary care clinics for potential mental health needs. Logistic regression was used to determine the predictors of past year marijuana use. RESULTS: A total of 11.5% of the sample reported using marijuana in the past year. Age, gender, other drug use, presence of alcohol use disorders, smoking status, depressive disorders, posttraumatic stress disorder, anxiety disorders, and psychotic symptoms, individually, were associated with the patients' use of marijuana during the past year. When controlling for age, race, and gender in a logistic regression analyses, only other drug use, alcohol use disorder, and smoking status were linked to past year marijuana use. Patients were 5.4 (95% confidence interval [CI] 4.3-6.7) times more likely to have used marijuana during the past year if they used another illicit drug during the past year. Those with alcohol use disorder diagnosis or current smokers were 2.3 (95% CI 1.9-2.8) and 1.5 times (95% CI 1.3-1.7), respectively, more likely to have used marijuana during the past year. Receiver operating characteristic curve (area under curve = 0.79) represents good sensitivity and specificity of the model, correctly classifying 88.4% of the past year marijuana users. CONCLUSION: Identifying patients at high risk for cannabis use may facilitate targeted screening and provision of interventions in primary care. Patients who screen positive for cigarette use, alcohol abuse or dependence, or have evidence of other illicit drug use could be considered for cannabis screening.

The association between chronic pain and prescription drug abuse in Veterans.

OBJECTIVE: We sought to investigate the association between chronic pain and self-reported prescription drug abuse in a large cohort of patients referred from primary care for a behavioral health assessment. DESIGN: We performed a cross-sectional analysis of responses to a telephone assessment administered to patients referred for a behavioral health evaluation between April 25, 2005 and October 31, 2007. We conducted descriptive statistics and investigated multivariable associations. Multivariable analyses included age, gender, race, financial status, employment, current smoking, drinking problem, past-year illicit drug use, depression, and chronic pain. PATIENTS: Veterans referred from primary care (N = 6,377). RESULTS: Mean age of the sample was 56.5 years with a range of 19-97. The majority of respondents was white, unmarried, and was unemployed. Nearly 5% of the sample reported past 6-month prescription drug abuse. On multivariable analysis, younger age, possible depression (odds ratio [OR] 1.9; 1.3-2.8), probable depression (OR 2.4; 1.6-3.4), smoking (OR 1.4; 1.1-1.8), illicit drug use (OR 2.8; 2.2-3.7), and chronic pain (OR 1.9; 1.4-2.5) were associated with prescription drug abuse. CONCLUSIONS: We have identified specific variables associated with self-reported prescription drug abuse in primary care patients. Chronic pain is associated both with an indication for prescribing opioids and with abuse of prescription medications. Clinicians are encouraged to follow treatment algorithms when managing patients with chronic pain as a method for reducing misuse.

Family problems among recently returned military veterans referred for a mental health evaluation.

CONTEXT: Existing evidence suggests that military veterans with mental health disorders have poorer family functioning, although little research has focused on this topic. OBJECTIVE: To test whether psychiatric symptoms are associated with family reintegration problems in recently returned military veterans. DESIGN: Cross-sectional survey of a clinical population. Respondents who were referred to behavioral health evaluation from April 2006 through August 2007 were considered for the survey. SETTING: Philadelphia Veterans Affairs Medical Center, Pa. PARTICIPANTS: 199 military veterans who served in Iraq or Afghanistan after 2001 and were referred for behavioral health evaluation from primary care (mean age = 32.7 years, SD = 9.1). MAIN OUTCOME MEASURES: Measures included the Mini-International Neuropsychiatric Interview for psychiatric diagnoses, the 9-item Patient Health Questionnaire for depression diagnosis and severity, and screening measures of alcohol abuse and illicit substance use. A measure of military family readjustment problems and a screening measure of domestic abuse were developed for this study. RESULTS: Three fourths of the married/cohabiting veterans reported some type of family problem in the past week, such as feeling like a guest in their household (40.7%), reporting their children acting afraid or not being warm toward them (25.0%), or being unsure about their family role (37.2%). Among veterans with current or recently separated partners, 53.7% reported conflicts involving "shouting, pushing, or shoving," and 27.6% reported that this partner was "afraid of them." Depression and posttraumatic stress disorder symptoms were both associated with higher rates of family reintegration problems. CONCLUSIONS: Mental health problems may complicate veterans' readjustment and reintegration into family life. The findings suggest an opportunity to improve the treatment of psychiatric disorders by addressing family problems.

Screening, assessment, and management of depression in VA primary care clinics. The Behavioral Health Laboratory.

OBJECTIVES: The purpose of this project was to assess the utility and feasibility of a telephone-based systematic clinical assessment service, the Behavioral Health Laboratory (BHL), in the context of primary care. The BHL is a clinical service that provides primary care providers with an assessment and a summary of mental health and substance abuse (MH/SA) symptoms and provides treatment decision support, including triage to specialty MH/SA services. The BHL was implemented to assist in the evaluation of patients who screened positively for depression at an annual clinical appointment or who were identified through routine care. METHODS: Results from systematic screening of primary care patients were extracted during a period of 6 months prior to implementation of the BHL and after implementation of the BHL. Descriptive results of the 580 evaluations conducted during this time were available. RESULTS: Results suggest an association between the implementation of the BHL and an increase in the proportion of patients screened for depression in primary care. In addition, there was an increase in the proportion of patients who screened positively (2.8% vs 7.0%). The BHL was successful in providing a comprehensive assessment for 78% of those referred. Significant co-occurring mental illness and substance misuse were found among those assessed. CONCLUSIONS: Introducing the BHL into primary care was associated with an apparent change in clinical practice in primary care at the Philadelphia VA Medical Center. Not only were more patients identified, the broad-based approach of the BHL identified significant comorbidity with alcohol misuse, illicit drugs, and suicidal ideation, symptoms likely to have been missed in routine clinical practice. The BHL offers a practical, low-cost method of assessment, monitoring, and treatment planning for patients identified in primary care with MH/SA needs.

Gambling among older, primary-care patients: an important public health concern.

OBJECTIVES: Authors assessed the prevalence rate of gambling in a sample of elderly patients (over age 65) and evaluated the sociodemographic characteristics, health, cognitive status, and psychiatric comorbidities of elderly gamblers. METHODS: This study is a cross-sectional survey of gambling habits. A random sample of older adults with a scheduled primary-care clinic appointment was selected and screened with the General Health Questionnaire and questions about suicidality and alcohol use. Also, participants were asked about their gambling habits. RESULTS: Of 843 screened patients completing the gambling questionnaire, 69.6% reported that they had participated in at least one gambling activity in the last year. At-risk gamblers were defined as those who reported having bet more than $100 on a single bet and/or having bet more than they could afford to lose in the last year. Of those responding, 10.9% were identified as at-risk gamblers. The strongest predictors of at-risk gambling behavior were being a binge drinker, presence of current posttraumatic stress disorder symptoms, minority race/ethnicity, and being a VA clinic patient. Subjects with mild-to-moderate cognitive impairment were just as likely as those without impairment to gamble and to report at-risk gambling behavior. At-risk gambling behavior was not significantly associated with gender, current or past depressive symptoms, or cigarette smoking. CONCLUSIONS: These data raise a significant public health concern that gambling behaviors are prevalent in older adults. Additional research is needed to further define the determinants of gambling behavior in older adults and evaluate the social, health, and economic costs and benefits of gambling by older adults, especially among those groups determined to be at risk.

Design and sample characteristics of the PRISM-E multisite randomized trial to improve behavioral health care for the elderly.

OBJECTIVE: To describe the design of the Primary Care Research in Substance Abuse and Mental Health for Elderly (PRISM-E) study and baseline characteristics of the randomized primary care patients with mental health problems and at-risk alcohol use. METHOD: Adults aged 65 and older were screened at primary care clinics from 10 study sites throughout the United States. Those diagnosed for depression, anxiety, and/or at-risk alcohol consumption were randomized to either integrated or enhanced referral care. RESULTS: Of the 23,828 participants, 14% had a positive assessment for depressive and/or anxiety disorders, and 6% had at-risk alcohol consumption diagnoses. Among patients with mental health diagnoses, there was a higher preponderance of younger ages, women, and ethnic minorities. Among patients with at-risk drinking, there was a higher preponderance of younger ages, Whites, and men. DISCUSSION: These findings indicate the need for screening in primary care and for engaging older adults in treatment.