Identifying Witnessed Suicides in National Violent Death Reporting System Narratives.

There is increasing attention to suicides that occur in view of others, as these deaths can cause significant psychological impact on witnesses. This study illuminates characteristics of witnessed suicides and compares characteristics of these deaths to non-witnessed suicides. We develop a codable definition of what constitutes witnessed (vs. non-witnessed) suicide. Our data include a sample of 1200 suicide descriptions from the 2003-2017 National Violent Death Reporting System (NVDRS). We first developed criteria to identify probable cases of witnessed suicide. The coding scheme achieved 94.5% agreement and identified approximately 10% (n = 125) of suicides as witnessed. Next, we examined differences between witnessed and non-witnessed suicides in demographics, manner of death, and social/environmental factors using bivariate Chi-squared tests, multivariate logistic regression, and ANOVA. Witnessed suicide decedents were significantly more likely than non-witnessed suicide decedents to be male, younger, and members of a sexual minority, and to have died in living spaces by means of a firearm. Two thirds of witnesses were strangers to the decedents, while 23.2% were romantic partners or ex-partners of the decedents. Our coding method offers a reliable approach to identify witnessed suicides. While witnessed suicides are relatively infrequent, these deaths have profound impact on witnesses. Articulating the features of witnessed suicides may contribute to identifying potential risk mitigation strategies.

Gendered Patterns in Manifest and Latent Mental Health Indicators Among Suicide Decedents: 2003-2020 National Violent Death Reporting System (NVDRS).

Objectives. To investigate differences in the documentation of mental health symptomology between male and female suicide decedents in the 2003-2020 US National Violent Death Reporting System (NVDRS). Methods. Using information on 271 998 suicides in the 2003-2020 NVDRS, we evaluated precoded mental health-related variables and topic model-derived latent mental health themes in the law enforcement and coroner or medical examiner death narratives compiled by trained public health workers. Results. Public health records of male compared with female suicides were less likely to include notations of mental health conditions or treatment interventions. However, topic modeling of death summaries revealed that male suicide decedents were more likely to evidence several subclinical cognitive and emotional indicators of distress. Conclusions. Suicide death records vary by gender, both in recorded evidence for mental health conditions at time of death and in accompanying narratives describing proximal circumstances surrounding these deaths. Our findings hint that patterns of subclinical mental health changes among men might be less well captured in commonly used mental health indicators, suggesting that prevention efforts may benefit from measures that also target assessment of subclinical distress. (Am J Public Health. 2024;114(S3):S268-S277. https://doi.org/10.2105/AJPH.2023.307427).

Prevalence of prescription and illicit drugs in suicides by non-poisoning means in the National Violent Death Reporting System 2003-2017.

Background: Prescription and illicit drugs are important social environmental variables in many suicides regardless of their role as an immediate cause of death. Objectives: To investigate the presence of prescription and illicit drugs, either through mention in the death record or toxicology reports, among suicides attributed to nonpoisonous causes to identify patterns of risk. Methods: Using the 2003-2017 National Violent Death Reporting System (NVDRS), we examined the presence of prescription and illicit drugs among 143,175 suicides (119,563 males 23,612 females) due to firearms and suffocation/hanging. The presence of drugs (opioids, stimulants, benzodiazepines, muscle relaxants, and cannabis) was determined from toxicology reports and text searches of coroner/medical examiner and law enforcement summaries. We fit multivariable logistic regression models to estimate associations between drug class and suicide method adjusting for decedent characteristics. Results: Overall prescription and illicit drugs were present in 22% of firearm deaths and 28% of suffocation deaths. Among victims with toxicology reports, over 20% tested positive for benzodiazepines. Benzodiazepines were mentioned in 4% of firearm and 5% of suffocation suicides without toxicology testing. Stimulants were more likely to occur in suffocation than firearm deaths among victims with toxicology testing (aOR = 1.44, 95% CI: 1.33-1.56) and without toxicology testing (aOR = 1.61, 95% CI: 1.31-1.98). Conclusions: Benzodiazepines were most frequently identified in both toxicology reports and narratives of suicides by firearms or suffocation. Better distinction of the presence of prescription and illicit drugs in the environment versus apparent ingestion among non-poisoning suicides are needed to inform prevention approaches.

Integrating topic modeling and word embedding to characterize violent deaths.

SignificanceWe introduce an approach to identify latent topics in large-scale text data. Our approach integrates two prominent methods of computational text analysis: topic modeling and word embedding. We apply our approach to written narratives of violent death (e.g., suicides and homicides) in the National Violent Death Reporting System (NVDRS). Many of our topics reveal aspects of violent death not captured in existing classification schemes. We also extract gender bias in the topics themselves (e.g., a topic about long guns is particularly masculine). Our findings suggest new lines of research that could contribute to reducing suicides or homicides. Our methods are broadly applicable to text data and can unlock similar information in other administrative databases.

Machine Learning as a Model for Cultural Learning: Teaching an Algorithm What it Means to be Fat.

Public culture is a powerful source of cognitive socialization; for example, media language is full of meanings about body weight. Yet it remains unclear how individuals process meanings in public culture. We suggest that schema learning is a core mechanism by which public culture becomes personal culture. We propose that a burgeoning approach in computational text analysis - neural word embeddings - can be interpreted as a formal model for cultural learning. Embeddings allow us to empirically model schema learning and activation from natural language data. We illustrate our approach by extracting four lower-order schemas from news articles: the gender, moral, health, and class meanings of body weight. Using these lower-order schemas we quantify how words about body weight "fill in the blanks" about gender, morality, health, and class. Our findings reinforce ongoing concerns that machine-learning models (e.g., of natural language) can encode and reproduce harmful human biases.

Displayed Depression Symptoms on Facebook at Two Time Points: Content Analysis.

BACKGROUND: Depression is a prevalent and problematic mental disorder that often has an onset in adolescence. Previous studies have illustrated that depression disclosures on social media are common and may be linked to an individual's experiences of depression. However, most studies have examined depression displays on social media at a single time point. OBJECTIVE: This study aims to investigate displayed depression symptoms on Facebook at 2 developmental time points based on symptom type and gender. METHODS: Participants were recruited from an ongoing longitudinal cohort study. The content analysis of text-based Facebook data over 1 year was conducted at 2 time points: time 1 (adolescence; age 17-18 years) and time 2 (young adulthood; ages 20-22 years). Diagnostic criteria for depression were applied to each post to identify the displayed depression symptoms. Data were extracted verbatim. The analysis included nonparametric tests for comparisons. RESULTS: A total of 78 participants' Facebook profiles were examined, of which 40 (51%) were male. At time 1, 62% (48/78) of the adolescents had a Facebook profile, and 54% (26/78) displayed depression symptom references with an average of 9.4 (SD 3.1) references and 3.3 (SD 2.3) symptom types. Of the 78 participants, 15 (19%) females and 12 (15%) males displayed depression symptom references; these prevalence estimates were not significantly different by gender (P=.59). At time 2, 35 young adults displayed symptoms of depression with an average of 4.6 (SD 2.3) references and 2.4 (SD 1.3) symptom types. There were no differences in the prevalence of symptoms of depression displayed between males (n=19) and females (n=16; P=.63). CONCLUSIONS: This content analysis study within an ongoing cohort study illustrates the differences in depression displays on Facebook by developmental stage and symptom. This study contributes to a growing body of literature by showing that using social media to observe and understand depression during the emerging adult developmental period may be a valuable approach.

Aggression, Escalation, and Other Latent Themes in Legal Intervention Deaths of Non-Hispanic Black and White Men: Results From the 2003‒2017 National Violent Death Reporting System.

Objectives. To investigate racial/ethnic differences in legal intervention‒related deaths using state-of-the-art topic modeling of law enforcement and coroner text summaries drawn from the 2003-2017 US National Violent Death Reporting System (NVDRS). Methods. Employing advanced topic modeling, we identified 8 topics consistent with dangerousness in death incidents in the NVDRS death narratives written by public health workers (PHWs). Using logistic regression, we then evaluated racial/ethnic differences in PHW-coded variables and narrative topics among 4981 males killed by legal intervention, while adjusting for age, county-level characteristics, and year. Results. Black, as compared with White, decedents were younger and their deaths were less likely to include PHW-coded mental health or substance use histories, weapon use, or positive toxicology for alcohol or psychoactive drugs, but more likely to include "gangs-as-an-incident-precipitant" coding. Topic modeling revealed less frequent thematic representation of "physical aggression" or "escalation" but more of "gangs or criminal networks" among Black versus White decedents. Conclusions. While Black males were more likely to be victims of legal intervention deaths, PHW-coded variables in the NVDRS and death narratives suggest lower threat profiles among Black versus similar White decedents. The source of this greater risk remains undetermined.

Development and Testing of a 3-Item Screening Tool for Problematic Internet Use.

OBJECTIVE: To develop and validate the Problematic and Risky Internet Use Screening Scale (PRIUSS)-3 screening scale, a short scale to screen for Problematic Internet Use. STUDY DESIGN: This scale development study applied standard processes using separate samples for training and testing datasets. We recruited participants from schools and colleges in 6 states and 2 countries. We selected 3 initial versions of a PRIUSS-3 using correlation to the PRIUSS-18 score. We evaluated these 3 potential screening scales for conceptual coherence, factor loading, sensitivity, and specificity. We selected a 3-item screening tool and evaluated it in 2 separate testing sets using receiver operating curves. RESULTS: Our study sample included 1079 adolescents and young adults. The PRIUSS-3 included items addressing anxiety when away from the Internet, loss of motivation when on the Internet, and feelings of withdrawal when away from the Internet. This screening scale had a sensitivity of 100% and specificity of 69%. A score of ≥3 on the PRIUSS-3 was the threshold to follow up with the PRIUSS-18. DISCUSSION: Similar to other clinical screening tools, the PRIUSS-3 can be administered quickly in a clinical or research setting. Positive screens should be followed by administering the full PRIUSS-18. Given the pervasive presence of the Internet in youth's lives, screening and counseling for Problematic Internet Use can be facilitated by use of this validated screening tool.

#Proana: Pro-Eating Disorder Socialization on Twitter.

PURPOSE: Pro-eating disorder (ED) online movements support engagement with ED lifestyles and are associated with negative health consequences for adolescents with EDs. Twitter is a popular social media site among adolescents that provides a unique setting for Pro-ED content to be publicly exchanged. The purpose of this study was to investigate Pro-ED Twitter profiles' references to EDs and how their social connections (followers) reference EDs. METHODS: A purposeful sample of 45 Pro-ED profiles was selected from Twitter. Profile information, all tweets, and a random sample of 100 of their followers' profile information were collected for content analysis using the Twitter Application Programming Interface. A codebook based on ED screening guidelines was applied to evaluate ED references. For each Pro-ED profile, proportion of tweets with ED references and proportion of followers with ED references in their own profile were evaluated. RESULTS: In total, our 45 Pro-ED profiles generated 4,245 tweets for analysis. A median of 36.4% of profiles' tweets contained ED references. Pro-ED profiles had a median of 173 followers, and a median of 44.5% of followers had ED references. Pro-ED profiles with more tweets with ED references also tended to have more followers with ED references (ß = .37, p < .01). CONCLUSIONS: Findings suggest that profiles which self-identify as Pro-ED express disordered eating patterns through tweets and have an audience of followers, many of whom also reference ED in their own profiles. ED socialization on Twitter might provide social support, but in the Pro-ED context this activity might also reinforce an ED identity.

Evaluating College Students' Displayed Alcohol References on Facebook and Twitter.

PURPOSE: Current trends suggest that adolescents and young adults typically maintain a social media "portfolio" of several sites including Facebook and Twitter, but little is known regarding how an individual chooses to display risk behaviors across these different sites. The purpose of this study was to investigate college students' displayed alcohol references on both Facebook and Twitter. METHODS: Among a larger sample of college students from two universities, we identified participants who maintained both Facebook and Twitter profiles. Data collection included evaluation of 5 months of participants' Facebook and Twitter posts for alcohol references, number of social connections (i.e., friends or followers), and number of posts. Phone interviews assessed participants' frequency of Facebook and Twitter use and self-reported alcohol use. Analyses included Fisher's exact test, Wilcoxon matched pair sign test, Friedman rank-sum tests, and logistic regression. RESULTS: Of 112 eligible participants, 94 completed the study. Participants were more likely to display alcohol references on Facebook compared with those on Twitter (76% vs. 34%, p = .02). Participants reported more social connections on Facebook versus Twitter (average 801.2 friends vs. 189.4 followers, p < .001) and were more likely to report daily use of Facebook versus Twitter (94.6% vs. 50%, p < .001). Current alcohol use was predictive of both Facebook and Twitter displayed alcohol references, but mediators differed in each model. CONCLUSIONS: College students were more likely to display alcohol references on Facebook compared with those on Twitter. Understanding these patterns and predictors may inform prevention and intervention efforts directed at particular social media sites.

Using A Facebook Group As An Adjunct To A Pilot mHealth Physical Activity Intervention: A Mixed Methods Approach.

In the United States, most adolescents do not obtain the recommended amounts of physical activity for optimal health. Around 80% of adolescents own a mobile device, and social media is frequently used by adolescents on mobile devices. Few studies have examined the use of social media as part of an intervention to promote physical activity. The purpose of this study was to evaluate the use of a Facebook group as part of a mHealth physical activity intervention trial. Adolescents, ages 14-18 years, were recruited for a four week physical activity intervention using the FitBit Flex. Participants were also given the option to join a private Facebook group where they could interact and were given badges for fitness accomplishments. The research assistant moderator posted on the Facebook group an average of 25.3 times (SD=7.2). Post-intervention, participants completed a phone interview about their experience. Of 30 intervention participants (avg age 16.0 (SD=1.1), 60.0% female), 17 opted to join the Facebook group (avg age 16.3 (SD=1.2), 47.0% female) of which 10 completed a qualitative interview. Participants averaged 4.9 interactions (SD=8.7) on the Facebook group wall throughout the intervention. From the interview responses, major themes included enjoying the badge feature of the Facebook group and wanting more content and interaction. In conclusion, participants used and enjoyed having the Facebook group, particularly the badge feature of the group, as an adjunct to the physical activity intervention.