Large language models could change the future of behavioral healthcare: a proposal for responsible development and evaluation.

Large language models (LLMs) such as Open AI's GPT-4 (which power ChatGPT) and Google's Gemini, built on artificial intelligence, hold immense potential to support, augment, or even eventually automate psychotherapy. Enthusiasm about such applications is mounting in the field as well as industry. These developments promise to address insufficient mental healthcare system capacity and scale individual access to personalized treatments. However, clinical psychology is an uncommonly high stakes application domain for AI systems, as responsible and evidence-based therapy requires nuanced expertise. This paper provides a roadmap for the ambitious yet responsible application of clinical LLMs in psychotherapy. First, a technical overview of clinical LLMs is presented. Second, the stages of integration of LLMs into psychotherapy are discussed while highlighting parallels to the development of autonomous vehicle technology. Third, potential applications of LLMs in clinical care, training, and research are discussed, highlighting areas of risk given the complex nature of psychotherapy. Fourth, recommendations for the responsible development and evaluation of clinical LLMs are provided, which include centering clinical science, involving robust interdisciplinary collaboration, and attending to issues like assessment, risk detection, transparency, and bias. Lastly, a vision is outlined for how LLMs might enable a new generation of studies of evidence-based interventions at scale, and how these studies may challenge assumptions about psychotherapy.

Differential temporal utility of passively sensed smartphone features for depression and anxiety symptom prediction: a longitudinal cohort study.

While studies show links between smartphone data and affective symptoms, we lack clarity on the temporal scale, specificity (e.g., to depression vs. anxiety), and person-specific (vs. group-level) nature of these associations. We conducted a large-scale (n = 1013) smartphone-based passive sensing study to identify within- and between-person digital markers of depression and anxiety symptoms over time. Participants (74.6% female; M age = 40.9) downloaded the LifeSense app, which facilitated continuous passive data collection (e.g., GPS, app and device use, communication) across 16 weeks. Hierarchical linear regression models tested the within- and between-person associations of 2-week windows of passively sensed data with depression (PHQ-8) or generalized anxiety (GAD-7). We used a shifting window to understand the time scale at which sensed features relate to mental health symptoms, predicting symptoms 2 weeks in the future (distal prediction), 1 week in the future (medial prediction), and 0 weeks in the future (proximal prediction). Spending more time at home relative to one's average was an early signal of PHQ-8 severity (distal ß = 0.219, p = 0.012) and continued to relate to PHQ-8 at medial (ß = 0.198, p = 0.022) and proximal (ß = 0.183, p = 0.045) windows. In contrast, circadian movement was proximally related to (ß = -0.131, p = 0.035) but did not predict (distal ß = 0.034, p = 0.577; medial ß = -0.089, p = 0.138) PHQ-8. Distinct communication features (i.e., call/text or app-based messaging) related to PHQ-8 and GAD-7. Findings have implications for identifying novel treatment targets, personalizing digital mental health interventions, and enhancing traditional patient-provider interactions. Certain features (e.g., circadian movement) may represent correlates but not true prospective indicators of affective symptoms. Conversely, other features like home duration may be such early signals of intra-individual symptom change, indicating the potential utility of prophylactic intervention (e.g., behavioral activation) in response to person-specific increases in these signals.

Tapping into alcohol use during COVID: Drinking correlates among bartenders and servers.

The COVID pandemic placed a spotlight on alcohol use and the hardships of working within the food and beverage industry, with millions left jobless. Following previous studies that have found elevated rates of alcohol problems among bartenders and servers, here we studied the alcohol use of bartenders and servers who were employed during COVID. From February 12-June 16, 2021, in the midst of the U.S. COVID national emergency declaration, survey data from 1,010 employed bartender and servers were analyzed to quantify rates of excessive or hazardous drinking along with regression predictors of alcohol use as assessed by the 10-item Alcohol Use Disorders Identification Test (AUDIT). Findings indicate that more than 2 out of 5 (44%) people surveyed reported moderate or high rates of alcohol problem severity (i.e., AUDIT scores of 8 or higher)-a rate 4 to 6 times that of the heavy alcohol use rate reported pre- or mid-pandemic by adults within and outside the industry. Person-level factors (gender, substance use, mood) along with the drinking habits of one's core social group were significantly associated with alcohol use. Bartenders and servers reported surprisingly high rates of alcohol problem severity and experienced risk factors for hazardous drinking at multiple ecological levels. Being a highly vulnerable and understudied population, more studies on bartenders and servers are needed to assess and manage the true toll of alcohol consumption for industry employees.

Key language markers of depression on social media depend on race.

Depression has robust natural language correlates and can increasingly be measured in language using predictive models. However, despite evidence that language use varies as a function of individual demographic features (e.g., age, gender), previous work has not systematically examined whether and how depression's association with language varies by race. We examine how race moderates the relationship between language features (i.e., first-person pronouns and negative emotions) from social media posts and self-reported depression, in a matched sample of Black and White English speakers in the United States. Our findings reveal moderating effects of race: While depression severity predicts I-usage in White individuals, it does not in Black individuals. White individuals use more belongingness and self-deprecation-related negative emotions. Machine learning models trained on similar amounts of data to predict depression severity performed poorly when tested on Black individuals, even when they were trained exclusively using the language of Black individuals. In contrast, analogous models tested on White individuals performed relatively well. Our study reveals surprising race-based differences in the expression of depression in natural language and highlights the need to understand these effects better, especially before language-based models for detecting psychological phenomena are integrated into clinical practice.

A cross-cultural examination of temporal orientation through everyday language on social media.

Past research has shown that culture can form and shape our temporal orientation-the relative emphasis on the past, present, or future. However, there are mixed findings on how temporal orientations vary between North American and East Asian cultures due to the limitations of survey methodology and sampling. In this study, we applied an inductive approach and leveraged big data and natural language processing between two popular social media platforms-Twitter and Weibo-to assess the similarities and differences in temporal orientation in the United States of America and China, respectively. We first established predictive models from annotation data and used them to classify a larger set of English Twitter sentences (NTW = 1,549,136) and a larger set of Chinese Weibo sentences (NWB = 95,181) into four temporal catetories-past, future, atemporal present, and temporal present. Results show that there is no significant difference between Twitter and Weibo on past or future orientations; the large temporal orientation difference between North Americans and Chinese derives from their different prevailing focus on atemporal (e.g., facts, ideas) present (Twitter) or temporal present (e.g., the "here" and "now") (Weibo). Our findings contribute to the debate on cultural differences in temporal orientations with new perspectives following a new methodological approach. The study's implications call for a reevaluation of how temporal orientation is measured in cross-cultural studies, emphasizing the use of large-scale language data and acknowledging the atemporal present category. Understanding temporal orientations can guide effective cross-cultural communication strategies to tailor approaches for different audience based on temporal orientations, enhancing intercultural understanding and engagement.

Language-based EMA assessments help understand problematic alcohol consumption.

BACKGROUND: Unhealthy alcohol consumption is a severe public health problem. But low to moderate alcohol consumption is associated with high subjective well-being, possibly because alcohol is commonly consumed socially together with friends, who often are important for subjective well-being. Disentangling the health and social complexities of alcohol behavior has been difficult using traditional rating scales with cross-section designs. We aim to better understand these complexities by examining individuals' everyday affective subjective well-being language, in addition to rating scales, and via both between- and within-person designs across multiple weeks. METHOD: We used daily language and ecological momentary assessment on 908 US restaurant workers (12692 days) over two-week intervals. Participants were asked up to three times a day to "describe your current feelings", rate their emotions, and report their alcohol behavior in the past 24 hours, including if they were drinking alone or with others. RESULTS: Both between and within individuals, language-based subjective well-being predicted alcohol behavior more accurately than corresponding rating scales. Individuals self-reported being happier on days when drinking more, with language characteristic of these days predominantly describing socializing with friends. Between individuals (over several weeks), subjective well-being correlated much more negatively with drinking alone (r = -.29) than it did with total drinking (r = -.10). Aligned with this, people who drank more alone generally described their feelings as sad, stressed and anxious and drinking alone days related to nervous and annoyed language as well as a lower reported subjective well-being. CONCLUSIONS: Individuals' daily subjective well-being, as measured via language, in part, explained the social aspects of alcohol drinking. Further, being alone explained this relationship, such that drinking alone was associated with lower subjective well-being.

Characterizing empathy and compassion using computational linguistic analysis.

Many scholars have proposed that feeling what we believe others are feeling-often known as "empathy"-is essential for other-regarding sentiments and plays an important role in our moral lives. Caring for and about others (without necessarily sharing their feelings)-often known as "compassion"-is also frequently discussed as a relevant force for prosocial motivation and action. Here, we explore the relationship between empathy and compassion using the methods of computational linguistics. Analyses of 2,356,916 Facebook posts suggest that individuals (N = 2,781) high in empathy use different language than those high in compassion, after accounting for shared variance between these constructs. Empathic people, controlling for compassion, often use self-focused language and write about negative feelings, social isolation, and feeling overwhelmed. Compassionate people, controlling for empathy, often use other-focused language and write about positive feelings and social connections. In addition, high empathy without compassion is related to negative health outcomes, while high compassion without empathy is related to positive health outcomes, positive lifestyle choices, and charitable giving. Such findings favor an approach to moral motivation that is grounded in compassion rather than empathy. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Smartphone sensor data estimate alcohol craving in a cohort of patients with alcohol-associated liver disease and alcohol use disorder.

BACKGROUND: Sensors within smartphones, such as accelerometer and location, can describe longitudinal markers of behavior as represented through devices in a method called digital phenotyping. This study aimed to assess the feasibility of digital phenotyping for patients with alcohol-associated liver disease and alcohol use disorder, determine correlations between smartphone data and alcohol craving, and establish power assessment for future studies to prognosticate clinical outcomes. METHODS: A total of 24 individuals with alcohol-associated liver disease and alcohol use disorder were instructed to download the AWARE application to collect continuous sensor data and complete daily ecological momentary assessments on alcohol craving and mood for up to 30 days. Data from sensor streams were processed into features like accelerometer magnitude, number of calls, and location entropy, which were used for statistical analysis. We used repeated measures correlation for longitudinal data to evaluate associations between sensors and ecological momentary assessments and standard Pearson correlation to evaluate within-individual relationships between sensors and craving. RESULTS: Alcohol craving significantly correlated with mood obtained from ecological momentary assessments. Across all sensors, features associated with craving were also significantly correlated with all moods (eg, loneliness and stress) except boredom. Individual-level analysis revealed significant relationships between craving and features of location entropy and average accelerometer magnitude. CONCLUSIONS: Smartphone sensors may serve as markers for alcohol craving and mood in alcohol-associated liver disease and alcohol use disorder. Findings suggest that location-based and accelerometer-based features may be associated with alcohol craving. However, data missingness and low participant retention remain challenges. Future studies are needed for further digital phenotyping of relapse risk and progression of liver disease.

Filling in the white space: Spatial interpolation with Gaussian processes and social media data.

Full national coverage below the state level is difficult to attain through survey-based data collection. Even the largest survey-based data collections, such as the CDC's Behavioral Risk Factor Surveillance System or the Gallup-Healthways Well-being Index (both with more than 300,000 responses p.a.) only allow for the estimation of annual averages for about 260 out of roughly U.S. 3,000 counties when a threshold of 300 responses per county is used. Using a relatively high threshold of 300 responses gives substantially higher convergent validity-higher correlations with health variables-than lower thresholds but covers a reduced and biased sample of the population. We present principled methods to interpolate spatial estimates and show that including large-scale geotagged social media data can increase interpolation accuracy. In this work, we focus on Gallup-reported life satisfaction, a widely-used measure of subjective well-being. We use Gaussian Processes (GP), a formal Bayesian model, to interpolate life satisfaction, which we optimally combine with estimates from low-count data. We interpolate over several spaces (geographic and socioeconomic) and extend these evaluations to the space created by variables encoding language frequencies of approximately 6 million geotagged Twitter users. We find that Twitter language use can serve as a rough aggregate measure of socioeconomic and cultural similarity, and improves upon estimates derived from a wide variety of socioeconomic, demographic, and geographic similarity measures. We show that applying Gaussian Processes to the limited Gallup data allows us to generate estimates for a much larger number of counties while maintaining the same level of convergent validity with external criteria (i.e., N = 1,133 vs. 2,954 counties). This work suggests that spatial coverage of psychological variables can be reliably extended through Bayesian techniques while maintaining out-of-sample prediction accuracy and that Twitter language adds important information about cultural similarity over and above traditional socio-demographic and geographic similarity measures. Finally, to facilitate the adoption of these methods, we have also open-sourced an online tool that researchers can freely use to interpolate their data across geographies.

Predicting the Effect of Proton Beam Therapy Technology on Pulmonary Toxicities for Patients With Locally Advanced Lung Cancer Enrolled in the Proton Collaborative Group Prospective Clinical Trial.

PURPOSE: This study aimed to predict the probability of grade ≥2 pneumonitis or dyspnea within 12 months of receiving conventionally fractionated or mildly hypofractionated proton beam therapy for locally advanced lung cancer using machine learning. METHODS AND MATERIALS: Demographic and treatment characteristics were analyzed for 965 consecutive patients treated for lung cancer with conventionally fractionated or mildly hypofractionated (2.2-3 Gy/fraction) proton beam therapy across 12 institutions. Three machine learning models (gradient boosting, additive tree, and logistic regression with lasso regularization) were implemented to predict Common Terminology Criteria for Adverse Events version 4 grade ≥2 pulmonary toxicities using double 10-fold cross-validation for parameter hyper-tuning without leak of information. Balanced accuracy and area under the curve were calculated, and 95% confidence intervals were obtained using bootstrap sampling. RESULTS: The median age of the patients was 70 years (range, 20-97), and they had predominantly stage IIIA or IIIB disease. They received a median dose of 60 Gy in 2 Gy/fraction, and 46.4% received concurrent chemotherapy. In total, 250 (25.9%) had grade ≥2 pulmonary toxicity. The probability of pulmonary toxicity was 0.08 for patients treated with pencil beam scanning and 0.34 for those treated with other techniques (P = 8.97e-13). Use of abdominal compression and breath hold were highly significant predictors of less toxicity (P = 2.88e-08). Higher total radiation delivered dose (P = .0182) and higher average dose to the ipsilateral lung (P = .0035) increased the likelihood of pulmonary toxicities. The gradient boosting model performed the best of the models tested, and when demographic and dosimetric features were combined, the area under the curve and balanced accuracy were 0.75 ± 0.02 and 0.67 ± 0.02, respectively. After analyzing performance versus the number of data points used for training, we observed that accuracy was limited by the number of observations. CONCLUSIONS: In the largest analysis of prospectively enrolled patients with lung cancer assessing pulmonary toxicities from proton therapy to date, advanced machine learning methods revealed that pencil beam scanning, abdominal compression, and lower normal lung doses can lead to significantly lower probability of developing grade ≥2 pneumonitis or dyspnea.

A transdiagnostic, dimensional classification of anxiety shows improved parsimony and predictive noninferiority to DSM.

The current conceptualization of anxiety in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5)-which includes 11 anxiety disorders plus additional anxiety-related conditions-does not align with accumulating evidence that anxiety is transdiagnostic and dimensional in nature. Transdiagnostic dimensional anxiety models have been proposed, yet they measure anxiety at either a very broad (e.g., "anxiety") or very narrow (e.g., "performance anxiety") level, overlooking intermediate properties of anxiety that cut across DSM disorders. Using indicators from a well-validated semistructured interview of anxiety-related disorders, we constructed intermediate-level transdiagnostic dimensions representing the intensity, avoidance, pervasiveness, and onset of anxiety. We captured these content-agnostic dimensions in a sample representing varying levels and forms of anxiety (N = 268), including individuals with generalized anxiety disorder, social anxiety disorder, panic disorder, agoraphobia, specific phobia, separation anxiety disorder, posttraumatic stress disorder, and obsessive-compulsive disorder (n = 205) and individuals with no psychopathology (n = 63). In preregistered analyses, our dimensional anxiety model showed noninferiority to DSM-5 diagnoses in predicting concurrent and prospective measures of anxiety-related impairment, anxiety vulnerabilities, comorbid depression, and suicidal ideation. These results held regardless of whether the dimensions were combined into a single composite or retained as separate components. Our transdiagnostic dimensional model offers meaningful gains in parsimony over DSM, with no loss of predictive power. This project provides a methodological framework for the empirical evaluation of other transdiagnostic dimensional models of psychopathology that have been proposed as alternatives to the DSM. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

The online language of work-personal conflict.

With the blurring of boundaries in this digital age, there is increasing concern around work-personal conflict. Assessing and tracking work-personal conflict is critical as it not only affects individual workers but is also a vital measure among broader well-being and economic indices. This inductive study examines the extent to which work-personal conflict corresponds to individuals' language use on social media. We apply an open-vocabulary analysis to the posts of 2810 Facebook users who also completed a survey for an established work-personal conflict scale. It was found that the language-based model can predict personal-to-work conflict (r = 0.23) and work-to-personal conflict (r = 0.15) and provide important insights into such conflicts. Specifically, we found that high personal-to-work conflict was associated with netspeak and swearing, while low personal-to-work conflict was associated with language about work and positivity. We found that high work-to-personal conflict was associated with negative emotion and negative tone, while low work-to-personal conflict was associated with positive emotion and language about birthdays.

Specific associations of passively sensed smartphone data with future symptoms of avoidance, fear, and physiological distress in social anxiety.

Background: Prior literature links passively sensed information about a person's location, movement, and communication with social anxiety. These findings hold promise for identifying novel treatment targets, informing clinical care, and personalizing digital mental health interventions. However, social anxiety symptoms are heterogeneous; to identify more precise targets and tailor treatments, there is a need for personal sensing studies aimed at understanding differential predictors of the distinct subdomains of social anxiety. Our objective was to conduct a large-scale smartphone-based sensing study of fear, avoidance, and physiological symptoms in the context of trait social anxiety over time. Methods: Participants (n = 1013; 74.6 % female; M age = 40.9) downloaded the LifeSense app, which collected continuous passive data (e.g., GPS, communication, app and device use) over 16 weeks. We tested a series of multilevel linear regression models to understand within- and between-person associations of 2-week windows of passively sensed smartphone data with fear, avoidance, and physiological distress on the self-reported Social Phobia Inventory (SPIN). A shifting sensor lag was applied to examine how smartphone features related to SPIN subdomains 2 weeks in the future (distal prediction), 1 week in the future (medial prediction), and 0 weeks in the future (proximal prediction). Results: A decrease in time visiting novel places was a strong between-person predictor of social avoidance over time (distal ß = -0.886, p = .002; medial ß = -0.647, p = .029; proximal ß = -0.818, p = .007). Reductions in call- and text-based communications were associated with social avoidance at both the between- (distal ß = -0.882, p = .002; medial ß = -0.932, p = .001; proximal ß = -0.918, p = .001) and within- (distal ß = -0.191, p = .046; medial ß = -0.213, p = .028) person levels, as well as between-person fear of social situations (distal ß = -0.860, p < .001; medial ß = -0.892, p < .001; proximal ß = -0.886, p < .001) over time. There were fewer significant associations of sensed data with physiological distress. Across the three subscales, smartphone data explained 9-12 % of the variance in social anxiety. Conclusion: Findings have implications for understanding how social anxiety manifests in daily life, and for personalizing treatments. For example, a signal that someone is likely to begin avoiding social situations may suggest a need for alternative types of exposure-based interventions compared to a signal that someone is likely to begin experiencing increased physiological distress. Our results suggest that as a prophylactic means of targeting social avoidance, it may be helpful to deploy interventions involving social exposures in response to decreases in time spent visiting novel places.

Using artificial intelligence to assess personal qualities in college admissions.

Personal qualities like prosocial purpose and leadership predict important life outcomes, including college success. Unfortunately, the holistic assessment of personal qualities in college admissions is opaque and resource intensive. Can artificial intelligence (AI) advance the goals of holistic admissions? While cost-effective, AI has been criticized as a "black box" that may inadvertently penalize already disadvantaged subgroups when used in high-stakes settings. Here, we consider an AI approach to assessing personal qualities that aims to overcome these limitations. Research assistants and admissions officers first identified the presence/absence of seven personal qualities in n = 3131 applicant essays describing extracurricular and work experiences. Next, we fine-tuned pretrained language models with these ratings, which successfully reproduced human codes across demographic subgroups. Last, in a national sample (N = 309,594), computer-generated scores collectively demonstrated incremental validity for predicting 6-year college graduation. We discuss challenges and opportunities of AI for assessing personal qualities.

A deep learning approach to personality assessment: Generalizing across items and expanding the reach of survey-based research.

Traditional methods of personality assessment, and survey-based research in general, cannot make inferences about new items that have not been surveyed previously. This limits the amount of information that can be obtained from a given survey. In this article, we tackle this problem by leveraging recent advances in statistical natural language processing. Specifically, we extract "embedding" representations of questionnaire items from deep neural networks, trained on large-scale English language data. These embeddings allow us to construct a high-dimensional space of items, in which linguistically similar items are located near each other. We combine item embeddings with machine learning algorithms to extrapolate participant ratings of personality items to completely new items that have not been rated by any participants. The accuracy of our approach is on par with incentivized human judges given an identical task, indicating that it predicts ratings of new personality items as accurately as people do. Our approach is also capable of identifying psychological constructs associated with questionnaire items and can accurately cluster items into their constructs based only on their language content. Overall, our results show how representations of linguistic personality descriptors obtained from deep language models can be used to model and predict a large variety of traits, scales, and constructs. In doing so, they showcase a new scalable and cost-effective method for psychological measurement. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Depression and anxiety have distinct and overlapping language patterns: Results from a clinical interview.

Depression has been associated with heightened first-person singular pronoun use (I-usage; e.g., "I," "my") and negative emotion words. However, past research has relied on nonclinical samples and nonspecific depression measures, raising the question of whether these features are unique to depression vis-à-vis frequently co-occurring conditions, especially anxiety. Using structured questions about recent life changes or difficulties, we interviewed a sample of individuals with varying levels of depression and anxiety (N = 486), including individuals in a major depressive episode (n = 228) and/or diagnosed with generalized anxiety disorder (n = 273). Interviews were transcribed to provide a natural language sample. Analyses isolated language features associated with gold standard, clinician-rated measures of depression and anxiety. Many language features associated with depression were in fact shared between depression and anxiety. Language markers with relative specificity to depression included I-usage, sadness, and decreased positive emotion, while negations (e.g., "not," "no"), negative emotion, and several emotional language markers (e.g., anxiety, stress, depression) were relatively specific to anxiety. Several of these results were replicated using a self-report measure designed to disentangle components of depression and anxiety. We next built machine learning models to detect severity of common and specific depression and anxiety using only interview language. Individuals' speech characteristics during this brief interview predicted their depression and anxiety severity, beyond other clinical and demographic variables. Depression and anxiety have partially distinct patterns of expression in spoken language. Monitoring of depression and anxiety severity via language can augment traditional assessment modalities and aid in early detection. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Analyzing text message linguistic features: Do people with depression communicate differently with their close and non-close contacts?

BACKGROUND: Relatively little is known about how communication changes as a function of depression severity and interpersonal closeness. We examined the linguistic features of outgoing text messages among individuals with depression and their close- and non-close contacts. METHODS: 419 participants were included in this 16-week-long observational study. Participants regularly completed the PHQ-8 and rated subjective closeness to their contacts. Text messages were processed to count frequencies of word usage in the LIWC 2015 libraries. A linear mixed modeling approach was used to estimate linguistic feature scores of outgoing text messages. RESULTS: Regardless of closeness, people with higher PHQ-8 scores tended to use more differentiation words. When texting with close contacts, individuals with higher PHQ-8 scores used more first-person singular, filler, sexual, anger, and negative emotion words. When texting with non-close contacts these participants used more conjunctions, tentative, and sadness-related words and fewer first-person plural words. CONCLUSION: Word classes used in text messages, when combined with symptom severity and subjective social closeness data, may be indicative of underlying interpersonal processes. These data may hold promise as potential treatment targets to address interpersonal drivers of depression.

Predicting U.S. county opioid poisoning mortality from multi-modal social media and psychological self-report data.

Opioid poisoning mortality is a substantial public health crisis in the United States, with opioids involved in approximately 75% of the nearly 1 million drug related deaths since 1999. Research suggests that the epidemic is driven by both over-prescribing and social and psychological determinants such as economic stability, hopelessness, and isolation. Hindering this research is a lack of measurements of these social and psychological constructs at fine-grained spatial and temporal resolutions. To address this issue, we use a multi-modal data set consisting of natural language from Twitter, psychometric self-reports of depression and well-being, and traditional area-based measures of socio-demographics and health-related risk factors. Unlike previous work using social media data, we do not rely on opioid or substance related keywords to track community poisonings. Instead, we leverage a large, open vocabulary of thousands of words in order to fully characterize communities suffering from opioid poisoning, using a sample of 1.5 billion tweets from 6 million U.S. county mapped Twitter users. Results show that Twitter language predicted opioid poisoning mortality better than factors relating to socio-demographics, access to healthcare, physical pain, and psychological well-being. Additionally, risk factors revealed by the Twitter language analysis included negative emotions, discussions of long work hours, and boredom, whereas protective factors included resilience, travel/leisure, and positive emotions, dovetailing with results from the psychometric self-report data. The results show that natural language from public social media can be used as a surveillance tool for both predicting community opioid poisonings and understanding the dynamic social and psychological nature of the epidemic.

Comparison of wellbeing structures based on survey responses and social media language: A network analysis.

Wellbeing is predominantly measured through surveys but is increasingly measured by analysing individuals' language on social media platforms using social media text mining (SMTM). To investigate whether the structure of wellbeing is similar across both data collection methods, we compared networks derived from survey items and social media language features collected from the same participants. The dataset was split into an independent exploration (n = 1169) and a final subset (n = 1000). After estimating exploration networks, redundant survey items and language topics were eliminated. Final networks were then estimated using exploratory graph analysis (EGA). The networks of survey items and those from language topics were similar, both consisting of five wellbeing dimensions. The dimensions in the survey- and SMTM-based assessment of wellbeing showed convergent structures congruent with theories of wellbeing. Specific dimensions found in each network reflected the unique aspects of each type of data (survey and social media language). Networks derived from both language features and survey items show similar structures. Survey and SMTM methods may provide complementary methods to understand differences in human wellbeing.

A daily diary study into the effects on mental health of COVID-19 pandemic-related behaviors.

BACKGROUND: Recommendations for promoting mental health during the COVID-19 pandemic include maintaining social contact, through virtual rather than physical contact, moderating substance/alcohol use, and limiting news and media exposure. We seek to understand if these pandemic-related behaviors impact subsequent mental health. METHODS: Daily online survey data were collected on adults during May/June 2020. Measures were of daily physical and virtual (online) contact with others; substance and media use; and indices of psychological striving, struggling and COVID-related worry. Using random-intercept cross-lagged panel analysis, dynamic within-person cross-lagged effects were separated from more static individual differences. RESULTS: In total, 1148 participants completed daily surveys [657 (57.2%) females, 484 (42.1%) males; mean age 40.6 (s.d. 12.4) years]. Daily increases in news consumed increased COVID-related worrying the next day [cross-lagged estimate = 0.034 (95% CI 0.018-0.049), FDR-adjusted p = 0.00005] and vice versa [0.03 (0.012-0.048), FDR-adjusted p = 0.0017]. Increased media consumption also exacerbated subsequent psychological struggling [0.064 (0.03-0.098), FDR-adjusted p = 0.0005]. There were no significant cross-lagged effects of daily changes in social distancing or virtual contact on later mental health. CONCLUSIONS: We delineate a cycle wherein a daily increase in media consumption results in a subsequent increase in COVID-related worries, which in turn increases daily media consumption. Moreover, the adverse impact of news extended to broader measures of psychological struggling. A similar dynamic did not unfold between the daily amount of physical or virtual contact and subsequent mental health. Findings are consistent with current recommendations to moderate news and media consumption in order to promote mental health.