Using machine learning with intensive longitudinal data to predict depression and suicidal ideation among medical interns over time.

BACKGROUND: Use of intensive longitudinal methods (e.g. ecological momentary assessment, passive sensing) and machine learning (ML) models to predict risk for depression and suicide has increased in recent years. However, these studies often vary considerably in length, ML methods used, and sources of data. The present study examined predictive accuracy for depression and suicidal ideation (SI) as a function of time, comparing different combinations of ML methods and data sources. METHODS: Participants were 2459 first-year training physicians (55.1% female; 52.5% White) who were provided with Fitbit wearable devices and assessed daily for mood. Linear [elastic net regression (ENR)] and non-linear (random forest) ML algorithms were used to predict depression and SI at the first-quarter follow-up assessment, using two sets of variables (daily mood features only, daily mood features + passive-sensing features). To assess accuracy over time, models were estimated iteratively for each of the first 92 days of internship, using data available up to that point in time. RESULTS: ENRs using only the daily mood features generally had the best accuracy for predicting mental health outcomes, and predictive accuracy within 1 standard error of the full 92 day models was attained by weeks 7-8. Depression at 92 days could be predicted accurately (area under the curve >0.70) after only 14 days of data collection. CONCLUSIONS: Simpler ML methods may outperform more complex methods until passive-sensing features become better specified. For intensive longitudinal studies, there may be limited predictive value in collecting data for more than 2 months.

Does coping reduce suicidal urges in everyday life? Evidence from a daily diary study of adolescent inpatients.

BACKGROUND: Youth suicide has been increasing at an alarming rate. Identifying how youth at risk for suicide cope with daily distress and suicidal thoughts could inform prevention and intervention efforts. We investigated the relationship between previous-day coping and next-day suicidal urge intensity in a high-risk adolescent sample for a 4-week period. We also investigated the influence of adolescents' average coping levels, over 4 weeks, on daily severity of suicidal urges. METHODS: A total of 78 adolescents completed daily diaries after psychiatric hospitalization (n = 1621 observations). Each day, adolescents reported their use of specific coping strategies, overall coping helpfulness, and intensity of suicidal urges. RESULTS: Greater professional support seeking from providers/crisis lines and perceptions of coping helpfulness on the previous day were associated with lower next-day suicidal urges. Adolescents who reported greater average use of cognitive strategies, personal support seeking from family/friends, and higher average perceptions of coping helpfulness, relative to others, had lower daily suicidal urges. Noncognitive strategy use was not related to daily suicidal urge intensity. CONCLUSION: Findings point to the benefit of intervention efforts focusing on strengthening personal and professional supportive relationships, assisting youth with developing a broader coping repertoire, and working with adolescents to identify strategies they perceive to be helpful.

Suicidality and Alcohol Use as Predictors of Future Suicidal Behavior in College Students.

AIMS: There has been a marked increase in suicide fatalities among college-age students in recent years. Moreover, heavy alcohol use, a well-known risk factor for suicide, is present on most campuses. Yet, no prospective studies have examined alcohol use patterns among college students as predictors of suicidal behaviors. METHODS: Online of 40,335 students at four universities took place at the beginning of four academic years, 2015-2018. Of these, 2296 met criteria for an increased risk of suicidal behavior and completed 1- and/or 6-month follow-up evaluation(s). Baseline assessments included the Alcohol Use Disorders Identification Test to quantify alcohol consumption and resulting problems, and measures of depression, suicidal ideation and suicidal behavior. RESULTS: Suicide attempts during follow-up were reported by 35 (1.5%) of high-risk students. Regression analyses indicated that baseline severity of alcohol use consequences, but not amount of alcohol consumption, was associated with greater odds of a follow-up suicide attempt after controlling for baseline suicidal ideation, functional impairment and history of suicide attempts. CONCLUSIONS: Among college students at elevated risk for suicide, the severity of alcohol-related consequences was a significant predictor of future suicide attempts. Alcohol consumption was not a significant predictor, suggesting that the amount students drink is less of a concern for suicidal behavior than are the problems (e.g. failing to meet expectations, experiencing blackouts) associated with drinking.

Short-term change and prediction of suicidal ideation among adolescents: a daily diary study following psychiatric hospitalization.

BACKGROUND: Our understanding of suicidal ideation (SI) and its risk precursors is largely informed by studies spanning over wide time intervals (weeks, months, years). Little is known about SI as it occurs in daily lives of individuals at risk for suicide, the extent to which suicidal thoughts are dynamic over short periods of time, and the degree to which theoretically informed risk factors predict near-term SI. METHODS: Thirty-four adolescents hospitalized due to last-month suicide attempt and/or last-week SI (76% female; ages 13-17) responded to daily surveys sent to their cell phones for four consecutive weeks after discharge (n = 652 observations). RESULTS: There was notable variability in day-to-day SI, with half of ideation ratings changing at least one within-person standard deviation from one day to the next. Results of mixed effects models revealed concurrent (same-day), but not short-term prospective (next-day), associations between SI (frequency, duration, urge) and well-established predictors (connectedness, burdensomeness, hopelessness). However, synergistic effects of low connectedness with either high burdensomeness or high hopelessness were reliably associated with more severe same- and next-day suicidal ideation. CONCLUSIONS: This study adds to emerging literature indicating that suicidal thoughts fluctuate considerably among individuals at risk for suicide, further extending it by focusing on adolescents in the critical posthospitalization period. Fostering high-risk adolescents' sense of connectedness to others may be an especially promising intervention target. Frequent assessment of SI and its predictors, independently and in combination, could help identify promising predictors of short-term risk and meaningful intervention targets in high-risk teens.

Parental Self-Efficacy to Support Teens During a Suicidal Crisis and Future Adolescent Emergency Department Visits and Suicide Attempts.

This study of adolescents seeking emergency department (ED) services and their parents examined parents' self-efficacy beliefs to engage in suicide prevention activities, whether these beliefs varied based on teens' characteristics, and the extent to which they were associated with adolescents' suicide-related outcomes. Participants included 162 adolescents (57% female, 81.5% Caucasian), ages 13-17, and their parents. At index visit, parents rated their self-efficacy to engage in suicide prevention activities and their expectations regarding their teen's future suicide risk. Adolescents' ED visits for suicide-related concerns and suicide attempts were assessed 4 months later. Parents endorsed high self-efficacy to engage in most suicide prevention activities. At the same time, they endorsed considerable doubt in being able to keep their child safe if the teen has thoughts of suicide and in their child not attempting suicide in the future. Parents whose teens experienced follow-up suicide-related outcomes endorsed, at clinically meaningful effect sizes, lower self-efficacy for recognizing suicide warning signs, for obtaining the teen's commitment to refrain from suicide, and for encouraging their teen to cope, as well as lower confidence that their teen will not attempt suicide; self-efficacy to recognize warning signs was at trend level. Despite endorsing high self-efficacy for the majority of suicide prevention activities, parents of high-risk teens expressed less confidence in their capacity to influence their teen's suicidal behavior, which could undermine parents' effort to implement these strategies. The relationship between parental self-efficacy and youth suicide-related outcomes points to its potential value in guiding clinical decision making and interventions.

Suicide Risk Screening in Healthcare Settings: Identifying Males and Females at Risk.

Suicide is the tenth leading cause of death in the United States, accounting for more than 42,000 deaths in 2014. Although this tragedy cuts across groups defined by age, gender, race/ethnicity, and geographic location, it is striking that nearly four times as many males as females die by suicide in the U.S. We describe the current regulations and recommendations for suicide risk screening in healthcare systems and also describe the aspirational goal of "Zero Suicide," put forth by the National Action Alliance for Suicide Prevention. We then provide information about suicide risk screening tools and steps to take when a patient screens positive for suicide risk. Given the substantially higher suicide rate among males than females, we argue that it is important to consider how we could optimize suicide risk screening strategies to identify males at risk and females at risk. Further research is needed to accomplish this goal. It is recommended that we consider multi-factorial suicide risk screens that incorporate risk factors known to be particularly important for males as well as computerized, adaptive screens that are tailored for the specific risk considerations of the individual patient, male or female. These strategies are not mutually exclusive. Finally, universal suicide risk screening in healthcare settings, especially primary care, specialty medical care, and emergency department settings, is recommended.

Connectedness to family, school, peers, and community in socially vulnerable adolescents.

Youth who feel connected to people and institutions in their communities may be buffered from other risk factors in their lives. As a result, increasing connectedness has been recommended as a prevention strategy. In this study, we examined connectedness among 224 youth (ages 12-15), recruited from an urban medical emergency department, who were at elevated risk due to bullying perpetration or victimization, or low social connectedness. Regression analyses examined multiple domains of connectedness (family, school, peer, community) in relation to adjustment. Youth who felt more connected to parents reported lower levels of depressive symptoms, suicidal ideation, non-suicidal self-injury, and conduct problems, higher self-esteem and more adaptive use of free time. Youth who felt more connected to their school reported lower levels of depressive symptoms, suicidal ideation, social anxiety, and sexual activity, as well as higher levels of self-esteem and more adaptive use of free time. Community connectedness was associated with less social anxiety but more sexual activity, and peer connectedness was not related to youth adjustment in this unique sample. Findings suggest that family and school connectedness may buffer youth on a trajectory of risk, and may therefore be important potential targets for early intervention services.

SELF-RATED EXPECTATIONS OF SUICIDAL BEHAVIOR PREDICT FUTURE SUICIDE ATTEMPTS AMONG ADOLESCENT AND YOUNG ADULT PSYCHIATRIC EMERGENCY PATIENTS.

BACKGROUND: This study's purpose was to examine the predictive validity and clinical utility of a brief measure assessing youths' own expectations of their future risk of suicidal behavior, administered in a psychiatric emergency (PE) department; and determine if youths' ratings improve upon a clinician-administered assessment of suicidal ideation severity. The outcome was suicide attempts up to 18 months later. METHODS: In this medical record review study, 340 consecutively presenting youths (ages 13-24) seeking PE services over a 7-month period were included. Subsequent PE visits and suicide attempts were retrospectively tracked for up to 18 months. The 3-item scale assessing patients' perception of their own suicidal behavior risk and the clinician-administered ideation severity scale were used routinely at the study site. RESULTS: Cox regression results showed that youths' expectations of suicidal behavior were independently associated with increased risk of suicide attempts, even after adjusting for key covariates. Results were not moderated by sex, suicide attempt history, or age. Receiver-operating characteristic (ROC) analyses indicated that self-assessed expectations of risk improved the predictive accuracy of the clinician-administered suicidal ideation measure. CONCLUSIONS: Youths' ratings indicative of lower confidence in maintaining safety uniquely predicted follow-up attempts and provided incremental validity over and above the clinician-administered assessment and improved its accuracy, suggesting their potential for augmenting suicide risk formulation. Assessing youths' own perceptions of suicide risk appears to be clinically useful, feasible to implement in PE settings, and, if replicated, promising for improving identification of youth at risk for suicidal behavior.

Predicting Future Suicide Attempts Among Adolescent and Emerging Adult Psychiatric Emergency Patients.

The purpose of this study was to longitudinally examine specific characteristics of suicidal ideation in combination with histories of suicide attempts and non-suicidal self-injury (NSSI) to best evaluate risk for a future attempt among high-risk adolescents and emerging adults. Participants in this retrospective medical record review study were 473 (53% female; 69% Caucasian) consecutive patients, ages 15 to 24 years (M=19.4 years) who presented for psychiatric emergency services during a 9-month period. These patients' medical records, including a clinician-administered Columbia-Suicide Severity Rating Scale, were coded at the index visit and at future visits occurring within the next 18 months. Logistic regression models were used to predict suicide attempts during this period. Socioeconomic status, suicidal ideation severity (i.e., intent, method), suicidal ideation intensity (i.e., frequency, controllability), a lifetime history of suicide attempt, and a lifetime history of NSSI were significant independent predictors of a future suicide attempt. Suicidal ideation added incremental validity to the prediction of future suicide attempts above and beyond the influence of a past suicide attempt, whereas a lifetime history of NSSI did not. Sex moderated the relationship between the duration of suicidal thoughts and future attempts (predictive for male patients but not female). Results suggest value in incorporating both past behaviors and current thoughts into suicide risk formulation. Furthermore, suicidal ideation duration warrants additional examination as a potential critical factor for screening assessments evaluating suicide risk among high-risk samples, particularly for male patients.

Columbia-suicide severity rating scale: predictive validity with adolescent psychiatric emergency patients.

OBJECTIVE: Despite the high prevalence of psychiatric emergency (PE) visits for attempted suicide and nonsuicidal self-injury (NSSI) among adolescents, we have limited information about assessment tools that are helpful in predicting subsequent risk for suicide attempts among adolescents in PE settings. This study examined the predictive validity of a highly promising instrument, the Columbia-Suicide Severity Rating Scale (C-SSRS). METHOD: Participants were 178 adolescents (44.4% male; ages 13-17 years) seeking PE services. The C-SSRS interview and selected medical chart data were collected for the index visit and subsequent visits during a 1-year follow-up. RESULTS: A suicide risk concern was the most common chief complaint (50.6%) in this sample, and nearly one third of the adolescents (30.4%) reported a lifetime history of suicide attempt at index visit. Sixty-two adolescents (34.8%) had at least one return PE visit during follow-up. Lifetime history of NSSI predicted both return PE visits and a suicide attempt at return visit. The C-SSRS intensity scale score was a significant predictor of a suicide attempt at return visit for both the full sample of adolescents and the subsample who reported suicidal ideation at their index visit. In this subsample, one specific item on the intensity scale, duration, was also a significant predictor of both a return PE visit and a suicide attempt at return visit. CONCLUSIONS: The C-SSRS intensity scale and NSSI had predictive validity for suicide attempts at return visit. Results also suggest that duration of adolescents' suicidal thoughts may be particularly important to risk for suicidal behavior, warranting further study.

Emotional suppression mediates the relation between adverse life events and adolescent suicide: implications for prevention.

Suicidal ideation substantially increases the odds of future suicide attempts, and suicide is the second leading cause of death among adolescents. A history of adverse life events has been linked with future suicidal ideation and attempts, although studies examining potential mediating variables have been scarce. One probable mediating mechanism is how the individual copes with adverse life events. For example, certain coping strategies appear to be more problematic than others in increasing future psychopathology, and emotional suppression in particular has been associated with poor mental health outcomes in adults and children. However, no studies to date have examined the potential mediating role of emotional suppression in the relation between adverse life events and suicidal thoughts/behavior in adolescence. The goal of the current study was to examine emotional suppression as a mediator in the relation between childhood adversity and future suicidal thoughts/behaviors in youth. A total of 625 participants, aged 14-19 years, seeking ER services were administered measures assessing adverse life events, coping strategies, suicidal ideation in the last 2 weeks, and suicide attempts in the last month. The results suggest that emotional suppression mediates the relation between adversity and both (1) suicidal thoughts and (2) suicide attempts above and beyond demographic variables and depressive symptoms. This study has important implications for interventions aimed at preventing suicidal thoughts and behavior in adolescents with histories of adversity.

A review of the peak-end rule in mental health contexts.

The peak-end rule, a memory heuristic in which the most emotionally salient part of an experience (i.e., peak) and conclusion of an experience (i.e., end) are weighted more heavily in summary evaluations, has been understudied in mental health contexts. The recent growth of intensive longitudinal methods has provided new opportunities for examining the peak-end rule in the retrospective recall of mental health symptoms, including measures often used in measurement-based care initiatives. Additionally, principles of the peak-end rule have significant potential to be applied to exposure-based therapy procedures. Additional research is needed to better understand the contexts in which, and persons for whom, the peak-end rule presents a greater risk of bias, to ultimately improve assessment strategies and clinical care.

Comparative Effectiveness of Three Digital Interventions for Adults Seeking Psychiatric Services: A Randomized Clinical Trial.

Importance: There is a substantial gap between demand for and availability of mental health services. Digital mental health interventions (DMHIs) are promising tools for bridging this gap, yet little is known about their comparative effectiveness. Objective: To assess whether patients randomized to a cognitive behavioral therapy (CBT)-based or mindfulness-based DMHI had greater improvements in mental health symptoms than patients randomized to the enhanced personalized feedback (EPF)-only DMHI. DESIGN,: SETTING, AND PARTICIPANTS This randomized clinical trial was conducted between May 13, 2020, and December 12, 2022, with follow-up at 6 weeks. Adult patients of outpatient psychiatry services across various clinics within the University of Michigan Health System with a scheduled or recent outpatient psychiatry appointment were recruited. Eligible patients were randomized to an intervention arm. All analyses followed the intent-to-treat principle. Interventions: Participants were randomized to 1 of 5 intervention arms: (1) EPF only; (2) Silvercloud only, a mobile application designed to deliver CBT strategies; (3) Silvercloud plus EPF; (4) Headspace only, a mobile application designed to train users in mindfulness practices; and (5) Headspace plus EPF. Main Outcomes and Measures: The primary outcome was change in depressive symptoms as measured by the Patient Health Questionnaire-9 (PHQ-9; score range: 0-27, with higher scores indicating greater depression symptoms). Secondary outcomes included changes in anxiety, suicidality, and substance use symptoms. Results: A total of 2079 participants (mean [SD] age, 36.8 [14.3] years; 1423 self-identified as women [68.4%]) completed the baseline survey. The baseline mean (SD) PHQ-9 score was 12.7 (6.4) and significantly decreased for all 5 intervention arms at 6 weeks (from -2.1 [95% CI, -2.6 to -1.7] to -2.9 [95% CI, -3.4 to -2.4]; n = 1885). The magnitude of change was not significantly different across the 5 arms (F4,1879 = 1.19; P = .31). Additionally, the groups did not differ in decrease in anxiety or substance use symptoms. However, the Headspace arms reported significantly greater improvements on a suicidality measure subscale compared with the Silvercloud arms (mean difference in mean change = 0.63; 95% CI, 0.20-1.06; P = .004). Conclusions and Relevance: This randomized clinical trial found decreases in depression and anxiety symptoms across all DMHIs and minimal evidence that specific applications were better than others. The findings suggest that DMHIs may provide support for patients during waiting list-related delays in care. Trial Registration: ClinicalTrials.gov Identifier: NCT04342494.

Management of suicidal risk in the emergency department: A clinical pathway using the computerized adaptive screen for suicidal youth.

Objective: Given the critical need for efficient and tailored suicide screening for youth presenting in the emergency department (ED), this study establishes validated screening score thresholds for the Computerized Adaptive Screen for Suicidal Youth (CASSY) and presents an example of a suicide risk classification pathway. Methods: Participants were primarily from the Study One derivation cohort of the Emergency Department Screen for Teens at Risk for Suicide (ED-STARS) enrolled in collaboration with Pediatric Emergency Care Applied Research Networks (PECARN). CASSY scores corresponded to the predicted probabilities of a suicide attempt in the next 3 months and risk thresholds were classified as minimal (<1%), low (1%-5%), moderate (5%-10%), and high (>10%). CASSY scores were compared to risk thresholds derived from clinical consensus and ED complaints and dispositions. CASSY risk thresholds were also examined as predictors of future suicide attempts in the Study Two validation cohort of ED-STARS. Results: A total of 1452 teens were enrolled with a median age of 15.2 years, 59.5% were female, 55.6% were White, 22% were Black, 22.3% were Latinx, and 42.8% received public assistance. The clinical consensus suicide risk groups were strongly associated with the CASSY-predicted risk thresholds. Suicide attempts in the Study Two cohort occurred at a frequency consistent with the CASSY-predicted thresholds. Conclusions: The CASSY can be a valuable tool in providing patient-specific risk probabilities for a suicide attempt at 3 months and tailor the threshold cutoffs based on the availability of local mental health resources. We give an example of a clinical risk pathway, which should include segmentation of the ED population by medical versus psychiatric chief complaint.

Intensive longitudinal assessment following index trauma to predict development of PTSD using machine learning.

There are significant challenges to identifying which individuals require intervention following exposure to trauma, and a need for strategies to identify and provide individuals at risk for developing PTSD with timely interventions. The present study seeks to identify a minimal set of trauma-related symptoms, assessed during the weeks following traumatic exposure, that can accurately predict PTSD. Participants were 2185 adults (Mean age=36.4 years; 64% women; 50% Black) presenting for emergency care following traumatic exposure. Participants received a 'flash survey' with 6-8 varying symptoms (from a pool of 26 trauma symptoms) several times per week for eight weeks following the trauma exposure (each symptom assessed ∼6 times). Features (mean, sd, last, worst, peak-end scores) from the repeatedly assessed symptoms were included as candidate variables in a CART machine learning analysis to develop a pragmatic predictive algorithm. PTSD (PCL-5 ≥38) was present for 669 (31%) participants at the 8-week follow-up. A classification tree with three splits, based on mean scores of nervousness, rehashing, and fatigue, predicted PTSD with an Area Under the Curve of 0.836. Findings suggest feasibility for a 3-item assessment protocol, delivered once per week, following traumatic exposure to assess and potentially facilitate follow-up care for those at risk.

Peak-end bias in retrospective recall of depressive symptoms on the PHQ-9.

Mental health care is built around patient recall and report of clinical symptoms. However, memories of events and experiences rely on cognitive heuristics that influence our recall. The peak-end bias, which refers to the tendency for the most intense and proximate aspects of an experience to disproportionately influence our memory, has been understudied in the context of mental health symptoms and may unduly influence self-reported symptoms, even in the context of standardized assessments. To determine whether the peak-end bias applies to the report of depressive symptoms on the standardized Patient Health Questionnaire-9 (PHQ-9) assessment, we compared two scores from daily mood assessments collected over a 2-week period from 4,322 medical interns (56% women; 60% non-Hispanic White). The peak-end-mood score, which averaged the single lowest and most recent mood scores over 2 weeks had a significantly stronger correlation with the PHQ-9 than the mean-mood score, which averaged all mood scores during the 2 weeks. Likelihood ratio tests and fit statistics provided further support that the peak-end-mood score was a significantly better predictor of depression than the mean-mood score. Results were consistent when limiting the sample to those with mild-to-severe depressive symptoms, and when only examining the two primary mood items as the dependent variable. These findings provide evidence for a modest peak-end recall bias for mood and depressive symptoms. There may be benefits to implementing intermittent assessment strategies to support clinical decision-making. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Alcohol intake in relation to suicidal ideation and behavior among university students.

ObjectiveHeavy alcohol use has been clearly linked to risk for suicidal behaviors and is also prevalent on many American college campuses. This report uses a large sample of college students to explore relationships between alcohol use, depressive symptoms, and suicidality. Methods: A brief suicide screen was completed by 40,335 university students at four pariticipating sites. Assessments quantified recent depressive symptoms, alcohol use, suicidal ideation, and suicide attempts. Results: Problems from alcohol use were consistently associated with suicidal thoughts and attempts in the previous month, and in the previous year, but the quantity of alcohol used was not. Alcohol related problems exerted effects on the likelihood of both suicide ideation and attempts beyond those explained by their relationship with depressive symptoms. Conclusions: Screens for individuals at increased risk for suicidal ideation and behavior should emphasize alcohol-related problems over quantification of alcohol intake.

The Cost of Universal Suicide Risk Screening for Adolescents in Emergency Departments.

Suicide is the second leading cause of death among adolescents. As nearly 20% of adolescents visit emergency departments (EDs) each year, EDs have an opportunity to identify previously unrecognized suicide risk. A novel Computerized Adaptive Screen for Suicidal Youth (CASSY) was shown in a multisite study to be predictive for suicide attempts within 3 months. This study uses site-specific data to estimate the cost of CASSY implementation with adolescents in general EDs. When used universally with all adolescents who are present and able to participate in the screening, the average cost was USD 5.77 per adolescent. For adolescents presenting with non-behavioral complaints, the average cost was USD 2.60 per adolescent. Costs were driven primarily by time and personnel required for the further evaluation of suicide risk for those screening positive. Thus, universal screening using the CASSY, at very low costs relative to the cost of an ED visit, can facilitate services needed for at-risk adolescents.

Potential Neuroprotective Effects of Dietary Omega-3 Fatty Acids on Stress in Alzheimer's Disease.

BACKGROUND: A large number of individual potentially modifiable factors are associated with risk for Alzheimer's disease (AD). However, less is known about the interactions between the individual factors. METHODS: In order to begin to examine the relationship between a pair of factors, we performed a pilot study, surveying patients with AD and controls for stress exposure and dietary omega-3 fatty acid intake to explore their relationship for risk of AD. RESULTS: For individuals with the greatest stress exposure, omega-3 fatty acid intake was significantly greater in healthy controls than in AD patients. There was no difference among those with low stress exposure. CONCLUSIONS: These initial results begin to suggest that omega-3 fatty acids may mitigate AD risk in the setting of greater stress exposure. This will need to be examined with larger populations and other pairs of risk factors to better understand these important relationships. Examining how individual risk factors interact will ultimately be important for learning how to optimally decrease the risk of AD.

Utilizing daily mood diaries and wearable sensor data to predict depression and suicidal ideation among medical interns.

BACKGROUND: Intensive longitudinal methods (ILMs) for collecting self-report (e.g., daily diaries, ecological momentary assessment) and passive data from smartphones and wearable sensors provide promising avenues for improved prediction of depression and suicidal ideation (SI). However, few studies have utilized ILMs to predict outcomes for at-risk, non-clinical populations in real-world settings. METHODS: Medical interns (N = 2881; 57 % female; 58 % White) were recruited from over 300 US residency programs. Interns completed a pre-internship assessment of depression, were given Fitbit wearable devices, and provided daily mood ratings (scale: 1-10) via mobile application during the study period. Three-step hierarchical logistic regressions were used to predict depression and SI at the end of the first quarter utilizing pre-internship predictors in step 1, Fitbit sleep/step features in step 2, and daily diary mood features in step 3. RESULTS: Passively collected Fitbit features related to sleep and steps had negligible predictive validity for depression, and no incremental predictive validity for SI. However, mean-level and variability in mood scores derived from daily diaries were significant independent predictors of depression and SI, and significantly improved model accuracy. LIMITATIONS: Work schedules for interns may result in sleep and activity patterns that differ from typical associations with depression or SI. The SI measure did not capture intent or severity. CONCLUSIONS: Mobile self-reporting of daily mood improved the prediction of depression and SI during a meaningful at-risk period under naturalistic conditions. Additional research is needed to guide the development of adaptive interventions among vulnerable populations.