Search

BACKGROUND Patients with substance use disorders (SUD) are at increased risk for symptom deterioration following treatment, with up to 60% resuming substance use the first year post treatment, and craving, together with neurocognitive and mental health variables play important roles in the understanding of the trajectories from abstinence to substance use. OBJECTIVE This study aims to improve our understanding of specific profiles of variables explaining SUD symptom deterioration. Specifically, how individual variability in mental health, neurocognitive functioning, and smartphone use is associated with craving and substance use in a young adult clinical population. METHODS In this pilot study, 26 patients with SUD were included 2 weeks prior to discharge from inpatient SUD treatment. Patients underwent baseline neuropsychological and mental health assessments and were equipped with smartwatches and downloaded an app for mobile sensor data collection on their smartphones. Every 2 days for up to 8 weeks the patients were administered mobile ecological momentary assessments (EMA) assessing substance use, craving, mental health, cognition, and a mobile go/no-go performance task. Repeated EMAs, as well as battery use sensor data was averaged across all days per subject and used as candidate input variables in models of craving intensity and relapse. RESULTS A total of 455 momentary assessments were completed. Using EMA and baseline data as candidate input variables, and craving and substance use as responses, automated model selection identified mean craving intensity as the most important predictor of substance use, with variability in self-reported impulsivity, mental health, and battery use as the significant predictors of craving intensity. CONCLUSIONS This study adds novelty by collecting high intensity data, for a considerable period of time, including both mental health, mobile sensor and mobile cognitive assessment data. The study also contributes to our knowledge about the most severe part of this clinical population in a vulnerable period during and after discharge from inpatient treatment. We confirmed the importance of temporal variability in executive dysfunction and mood in explaining variability in craving, and in addition that phone use may add to this understanding. Finally, we found that craving intensity is an important explanatory variable in relapse prediction. Our results inform further work towards precision medicine and just-in-time interventions with young adults with substance use disorders and may serve as decision support for clinicians evaluating discharge. CLINICALTRIAL

BACKGROUND Mental disorders are prevalent during adolescence. Among the digital phenotypes currently being developed to monitor mental health symptoms, typing behavior is one promising candidate. However, few studies have directly assessed associations between typing behavior and mental health symptom severity, and whether these relationships differs between genders. OBJECTIVE In a cross-sectional analysis of a large cohort, we tested whether various features of typing behavior derived from keystroke metadata were associated with mental health symptoms and whether these relationships differed between genders. METHODS A total of 934 adolescents from the Future Proofing study undertook 2 typing tasks on their smartphones through the Future Proofing app. Common keystroke timing and frequency features were extracted across tasks. Mental health symptoms were assessed using the Patient Health Questionnaire-Adolescent version, the Children’s Anxiety Scale-Short Form, the Distress Questionnaire 5, and the Insomnia Severity Index. Bivariate correlations were used to test whether keystroke features were associated with mental health symptoms. The false discovery rates of P values were adjusted to q values. Machine learning models were trained and tested using independent samples (ie, 80% train 20% test) to identify whether keystroke features could be combined to predict mental health symptoms. RESULTS Keystroke timing features showed a weak negative association with mental health symptoms across participants. When split by gender, females showed weak negative relationships between keystroke timing features and mental health symptoms, and weak positive relationships between keystroke frequency features and mental health symptoms. The opposite relationships were found for males (except for dwell). Machine learning models using keystroke features alone did not predict mental health symptoms. CONCLUSIONS Increased mental health symptoms are weakly associated with faster typing, with important gender differences. Keystroke metadata should be collected longitudinally and combined with other digital phenotypes to enhance their clinical relevance. CLINICALTRIAL Australian and New Zealand Clinical Trial Registry, ACTRN12619000855123; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=377664&isReview=true

Background Mental disorders are prevalent during adolescence. Among the digital phenotypes currently being developed to monitor mental health symptoms, typing behavior is one promising candidate. However, few studies have directly assessed associations between typing behavior and mental health symptom severity, and whether these relationships differs between genders. Objective In a cross-sectional analysis of a large cohort, we tested whether various features of typing behavior derived from keystroke metadata were associated with mental health symptoms and whether these relationships differed between genders. Methods A total of 934 adolescents from the Future Proofing study undertook 2 typing tasks on their smartphones through the Future Proofing app. Common keystroke timing and frequency features were extracted across tasks. Mental health symptoms were assessed using the Patient Health Questionnaire-Adolescent version, the Children’s Anxiety Scale-Short Form, the Distress Questionnaire 5, and the Insomnia Severity Index. Bivariate correlations were used to test whether keystroke features were associated with mental health symptoms. The false discovery rates of P values were adjusted to q values. Machine learning models were trained and tested using independent samples (ie, 80% train 20% test) to identify whether keystroke features could be combined to predict mental health symptoms. Results Keystroke timing features showed a weak negative association with mental health symptoms across participants. When split by gender, females showed weak negative relationships between keystroke timing features and mental health symptoms, and weak positive relationships between keystroke frequency features and mental health symptoms. The opposite relationships were found for males (except for dwell). Machine learning models using keystroke features alone did not predict mental health symptoms. Conclusions Increased mental health symptoms are weakly associated with faster typing, with important gender differences. Keystroke metadata should be collected longitudinally and combined with other digital phenotypes to enhance their clinical relevance. Trial Registration Australian and New Zealand Clinical Trial Registry, ACTRN12619000855123; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=377664&isReview=true

Background Mood disorders are burdensome illnesses that often go undetected and untreated. Sensor technologies within smartphones may provide an opportunity for identifying the early changes in circadian rhythm and social support/connectedness that signify the onset of a depressive or manic episode. Objective Using smartphone sensor data, this study investigated the relationship between circadian rhythm, which was determined by GPS data, and symptoms of mental health among a clinical sample of adults diagnosed with major depressive disorder or bipolar disorder. Methods A total of 121 participants were recruited from a clinical setting to take part in a 10-week observational study. Self-report questionnaires for mental health outcomes, social support, social connectedness, and quality of life were assessed at 6 time points throughout the study period. Participants consented to passively sharing their smartphone GPS data for the duration of the study. Circadian rhythm (ie, regularity of location changes in a 24-hour rhythm) was extracted from GPS mobility patterns at baseline. Results Although we found no association between circadian rhythm and mental health functioning at baseline, there was a positive association between circadian rhythm and the size of participants’ social support networks at baseline (r=0.22; P=.03; R2=0.049). In participants with bipolar disorder, circadian rhythm was associated with a change in anxiety from baseline; a higher circadian rhythm was associated with an increase in anxiety and a lower circadian rhythm was associated with a decrease in anxiety at time point 5. Conclusions Circadian rhythm, which was extracted from smartphone GPS data, was associated with social support and predicted changes in anxiety in a clinical sample of adults with mood disorders. Larger studies are required for further validations. However, smartphone sensing may have the potential to monitor early symptoms of mood disorders.

BACKGROUND Mood disorders are burdensome illnesses that often go undetected and untreated. Sensor technologies within smartphones may provide an opportunity for identifying the early changes in circadian rhythm and social support/connectedness that signify the onset of a depressive or manic episode. OBJECTIVE Using smartphone sensor data, this study investigated the relationship between circadian rhythm, which was determined by GPS data, and symptoms of mental health among a clinical sample of adults diagnosed with major depressive disorder or bipolar disorder. METHODS A total of 121 participants were recruited from a clinical setting to take part in a 10-week observational study. Self-report questionnaires for mental health outcomes, social support, social connectedness, and quality of life were assessed at 6 time points throughout the study period. Participants consented to passively sharing their smartphone GPS data for the duration of the study. Circadian rhythm (ie, regularity of location changes in a 24-hour rhythm) was extracted from GPS mobility patterns at baseline. RESULTS Although we found no association between circadian rhythm and mental health functioning at baseline, there was a positive association between circadian rhythm and the size of participants’ social support networks at baseline (r=0.22; P=.03; R2=0.049). In participants with bipolar disorder, circadian rhythm was associated with a change in anxiety from baseline; a higher circadian rhythm was associated with an increase in anxiety and a lower circadian rhythm was associated with a decrease in anxiety at time point 5. CONCLUSIONS Circadian rhythm, which was extracted from smartphone GPS data, was associated with social support and predicted changes in anxiety in a clinical sample of adults with mood disorders. Larger studies are required for further validations. However, smartphone sensing may have the potential to monitor early symptoms of mood disorders.

High trait impulsivity is thought to contribute to the sense of loss of control over eating and impulses to binge eat experienced by those with Binge Eating Disorder (BED). Lisdexamfetamine Dimesylate (LDX), a drug approved for treatment of moderate to severe BED, has been shown to decrease impulsive features of BED. However, the relationship between LDX-related reductions of binge eating (BE) episodes and impulsivity has not yet been explored. Forty-one adults aged 18-40 years with moderate to severe BED completed questionnaires and tasks assessing impulsivity at baseline and after 8 weeks of 50-70mg of LDX. Twenty age-matched healthy controls were also assessed at 2 timepoints for normative comparison. Data were analysed using linear mixed models. BED participants exhibited increased self-reported motor, non-planning, cognitive and food-related impulsivity relative to controls but no differences in objective task-based measures of impulsivity. Food-related and non-planning impulsivity were significantly reduced by LDX, but not to normative levels. Individuals with higher baseline levels of motor and non-planning impulsivity, and loss of control over eating scores experienced the greatest reduction in BE frequency after 8 weeks of LDX. Further, there were significant associations between the degree to which subjective loss of control over eating, non-planning impulsivity and BE eating frequency reduced after 8 weeks of LDX. These data suggest that specific subjective measures of impulsivity may be able to predict who will have the greatest benefit from LDX treatment, and that reductions in binge eating frequency may be moderated by concurrent reductions in non-planning impulsivity.

Objective: The trail making task is used to assess executive functioning in ADHD youth, yet has only been validated in adult populations. We compare the relative contributions of various cognitive measures to performance on a trail making task analog, the Switching of Attention (SoA) task, in typically-developing and ADHD adolescents. Method: Participants were 160 adolescents with ADHD from the International Study to Predict Optimized Treatment—in ADHD, assessed at pretreatment baseline and 6-week medicated follow-up, and 160 matched typically-developing peers. Attention, processing speed, working memory, impulsivity, and motor speed were assessed using a cognitive battery. Results: Processing speed and working memory significantly contributed to SoA performance in ADHD, regardless of medication status. While medicated, motor speed also underpinned the prediction of most task measures. For typically-developing adolescents, sustained attention and working memory contributed to SoA performance. Conclusion: Typically-developing, unmedicated and treated ADHD adolescents recruit different aspects of cognition during SoA completion.

Side effects to antidepressant medications are common and can impact the prognosis of successful treatment outcome in people with major depressive disorder (MDD). However, few studies have investigated the severity of side effects over the course of treatment and their association with treatment outcome. Here we assessed the severity of side effects and the impact of treatment type and anxiety symptoms over the course of treatment, as well as whether side effects were associated with treatment outcome. Participants were N = 1008 adults with a current diagnosis of single-episode or recurrent, nonpsychotic MDD. Participants were randomised to receive escitalopram, sertraline, or venlafaxine-extended release with equal probability and reassessed at 8 weeks regarding Hamilton Rating Scale Depression (HRSD17) and Quick Inventory of Depressive Symptomatology (QIDS-SR16) remission and response. Severity of side effects were assessed using the Frequency, Intensity, and Burden of Side Effects Rating (FIBSER) scale and assessed at day 4 and weeks 2, 4, 6, and 8. Frequency, intensity, and burden of side effects were greatest at week 2, then only frequency and intensity of side effects gradually decreased up to week 6. Treatment type and anxiety symptoms did not impact the severity of side effects. A greater burden—but not frequency or intensity—of side effects was associated with poorer treatment outcome and as early as 4 days post-treatment. Together, this work provides an informative mapping of the progression of side effects throughout the treatment course and their association with treatment outcome. Importantly, the burden of side effects that are present as early as 4 days post-treatment predicts poorer treatment outcome and should be monitored closely. iSPOT-D: Registry name: ClinicalTrials.gov. Registration number: NCT00693849.

Accumulating evidence indicates that psychological and neurophysiological processes interconnect and interact to activate the body’s stress system and to trigger and maintain functional somatic symptoms. This study used the Early Life Stress Questionnaire, Depression Anxiety Stress Scales and biological markers (heart rate, heart rate variability, skin conductance, C-reactive protein (CRP) titre, respiratory rate, and accuracy and reaction time in an emotion-face identification task), to examine childhood adversity, psychological distress and stress-system activation in 35 children and adolescents (23 girls and 12 boys, 9–17 years old) disabled by chronic pain (vs two groups of age- and sex-matched healthy controls). Patients reported more early-life stress ( U = 798.5, p = .026) and more psychological distress ( U = 978, p < .001). They showed activation of the autonomic system: elevated heart rate ( U = 862.5, p = .003), elevated electrodermal activity ( U = 804.5, p = .024) and lower heart rate variability in the time domain ( U = 380.5, p = .007) and frequency domain ( U = 409.5, p = .017). The group showed an upward shift of CRP titres (with 75th and 90th CRP percentiles of 4.5 and 10.5 mg/L, respectively), suggesting the activation of the immune–inflammatory system. Elevated CRP titres were associated with elevated heart rate ( p = .028). There were no differences in respiratory rate or in accuracy and reaction time in the emotion-face identification task. The results indicate that interventions for children and adolescents with chronic pain need a multidisciplinary mind–body approach that concurrently addresses psychological distress, physical impairment and stress-system dysregulation.

BackgroundMajor depressive disorder (MDD) commonly co-occurs with clinically significant levels of anxiety. However, anxiety symptoms are varied and have been inconsistently associated with clinical, functional, and antidepressant treatment outcomes. We aimed to identify and characterise dimensions of anxiety in people with MDD and their use in predicting antidepressant treatment outcome.Method1008 adults with a current diagnosis of single-episode or recurrent, nonpsychotic, MDD were assessed at baseline on clinical features and cognitive/physiological functioning. Participants were then randomised to one of three commonly prescribed antidepressants and reassessed at 8 weeks regarding symptom change, as well as remission and response, on the 17-item Hamilton Rating Scale Depression (HRSD17) and the 16-item Quick Inventory of Depressive Symptomatology (QIDS-SR16). Exploratory factor analysis was used on items from scales assessing anxiety symptoms, and resulting factors were assessed against clinical features and cognitive/physiological functioning. Factors were also assessed on their ability to predict treatment outcome.ResultsThree factors emerged relating to stress, cognitive anxiety, and somatic anxiety. All factors showed high internal consistency, minimal cross-loadings, and unique clinical and functional profiles. Furthermore, only higher somatic anxiety was associated with poorer QIDS-SR16 remission, even after adjusting for covariates and multiple comparisons.ConclusionsAnxiety symptoms in people with MDD can be separated onto distinct factors that differentially respond to treatment outcome. Furthermore, these factors do not align with subscales of established measures of anxiety. Future research should consider cognitive and somatic symptoms of anxiety separately when assessing anxiety in MDD and their use in predicting treatment outcome.

Background Antidepressant efficacy in people with major depressive disorder remains modest, yet identifying treatment-predictive neurobiological markers may improve outcomes. While disruptions in functional connectivity within and between large-scale brain networks predict poorer treatment outcome, it is unclear whether higher trait neuroticism, which has been associated with generally poorer outcomes, contributes to these disruptions and to antidepressant-specific treatment outcomes. Here, we used whole-brain functional connectivity analysis to identify a neural connectomic signature of neuroticism and tested whether this signature predicted antidepressant treatment outcome. Methods Participants were 226 adults with major depressive disorder and 68 healthy control subjects who underwent functional magnetic resonance imaging and were assessed on clinical features at baseline. Participants with major depressive disorder were then randomized to 1 of 3 commonly prescribed antidepressants and after 8 weeks completed a second functional magnetic resonance imaging and were reassessed for depressive symptom remission/response. Baseline intrinsic functional connectivity between each pair of 436 brain regions was analyzed using network-based statistics to identify connectomic features associated with neuroticism. Features were then assessed on their ability to predict treatment outcome and whether they changed after 8 weeks of treatment. Results Higher baseline neuroticism was associated with greater connectivity within and between the salience, executive control, and somatomotor brain networks. Greater connectivity across these networks predicted poorer treatment outcome that was not mediated by baseline neuroticism, and connectivity strength decreased after antidepressant treatment. Conclusions Our findings demonstrate that neuroticism is associated with organization of intrinsic neural networks that predict treatment outcome, elucidating its biological underpinnings and opportunity for better treatment personalization.

Evidence from functional neuroimaging studies support neural differences between the Attention Deficit Hyperactivity Disorder (ADHD) presentation types. It remains unclear if these neural deficits also manifest at the structural level. We have previously shown that the ADHD combined, and ADHD inattentive types demonstrate differences in graph properties of structural covariance suggesting an underlying difference in neuroanatomical organization. The goal of this study was to examine and validate white matter brain organization between the two subtypes using both scalar and connectivity measures of brain white matter. We used both tract-based spatial statistical (TBSS) and tractography analyses with network-based Statistics (NBS) and graph-theoretical analyses in a cohort of 35 ADHD participants (aged 8-17 years) defined using DSM-IV criteria as combined (ADHD-C) type (n=19) or as predominantly inattentive (ADHD-I) type (n=16), and 28 matched neurotypical controls. We performed TBSS analyses on scalar measures of fractional anisotropy (FA), mean (MD), radial (RD), and axial (AD) diffusivity to assess differences in WM between ADHD types and controls. NBS and graph theoretical analysis of whole brain inter-regional tractography examined connectomic differences and brain network organization, respectively. None of the scalar measures significantly differed between ADHD types or relative to controls. Similarly, there were no tractography connectivity differences between the two subtypes and relative to controls using NBS. Global and regional graph measures were also similar between the groups. A single significant finding was observed for nodal degree between the ADHD-C and controls, in the right insula (corrected p=.029). Our result of no white matter differences between the subtypes is consistent with most previous findings. These findings together might suggest that the white matter structural architecture is largely similar between the DSM-based ADHD presentations.

Behavioural disturbances in attention deficit hyperactivity disorder (ADHD) are thought to be due to dysfunction of spatially distributed, interconnected neural systems. While there is a fast-growing literature on functional dysconnectivity in ADHD, far less is known about the structural architecture underpinning these disturbances and how it may contribute to ADHD symptomology and treatment prognosis. We applied graph theoretical analyses on diffusion MRI tractography data to produce quantitative measures of global network organisation and local efficiency of network nodes. Support vector machines (SVMs) were used for comparison of multivariate graph measures of 37 children and adolescents with ADHD relative to 26 age and gender matched typically developing children (TDC). We also explored associations between graph measures and functionally-relevant outcomes such as symptom severity and prediction of methylphenidate (MPH) treatment response. We found that multivariate patterns of reduced local efficiency, predominantly in subcortical regions (SC), were able to distinguish between ADHD and TDC groups with 76% accuracy. For treatment prognosis, higher global efficiency, higher local efficiency of the right supramarginal gyrus and multivariate patterns of increased local efficiency across multiple networks at baseline also predicted greater symptom reduction after 6 weeks of MPH treatment. Our findings demonstrate that graph measures of structural topology provide valuable diagnostic and prognostic markers of ADHD, which may aid in mechanistic understanding of this complex disorder.

Background Major Depressive Disorder (MDD), anxiety disorders, and high levels of anxious symptoms are associated with impaired cognitive functioning. However, little is known of how cognitive functioning is impaired in people with anxious depression. Here, we compared cognitive functioning between people with anxious depression, non-anxious depression, and healthy controls. We also tested whether anxious depression moderated the relationship between cognitive functioning and treatment outcome. Methods 1008 adults with MDD and 336 healthy controls completed IntegNeuro: a computerized cognitive functioning test battery. Participants were then randomised to one of three antidepressants and reassessed at 8 weeks using the 17-item Hamilton Depression Rating Scale (HRSD17) and the 16-Item Quick Inventory of Depressive Symptomatology-Self-Rated for remission and response. Syndromal anxious depression was defined as MDD with a comorbid anxiety disorder. HRSD anxious depression was defined as MDD with a comorbid HRSD17 anxiety/somatisation factor score ≥ 7. Results Syndromal anxious depression was associated with better psychomotor functioning and poorer working memory, cognitive flexibility and information processing speed compared to their non-anxious counterparts. HRSD anxious depression was associated with better psychomotor functioning compared to their non-anxious counterparts. Syndromal anxious depression moderated the relationship between verbal memory and treatment outcome. In people with syndromal anxious depression, poorer baseline verbal memory predicted poorer treatment outcome. Limitations As DSM-IV criteria was used, the DSM-5 anxious distress specifier characterisation of anxious depression could not be assessed Conclusions Syndromal anxious depression is characterised by impaired executive functions and moderates the relationship between verbal memory functioning and treatment outcome.

Background and Objectives: Chronically stressed individuals report deficits spanning cognitive and emotional functioning. However, limitations to clinical populations and measures of stress have impeded the generalisability and scope of results. This study investigated whether chronic stress predicted cognitive and emotional functioning, and whether these relationships differed between males and females, in a large representative sample of healthy participants. Design: Cross-sectional study. Method: 1883 healthy adults sampled from the Brain Resource International Database reported stress using the 21-item Depression Anxiety Stress Scales. Participants then completed a cognitive and emotional assessment battery (IntegNeuro), as well as questionnaires related to sleep, emotional functioning, and self-regulation. Results: In contrast to previously reported results, chronic stress did not predict cognitive functioning. However, higher stress predicted a greater negativity bias and poorer social skills, confirming previous research identifying these links. Conclusions: Cognitive deficits related to stress are absent in healthy participants when stress is measured using the 21-items Depression Anxiety Stress Scales. Identifying how chronic stress is associated with aspects of emotional functioning can lead to personalized interventions for individuals to better manage the negative outcomes resulting from stress.

Objective: Major depressive disorder commonly co-occurs with one or more anxiety disorders or with clinically significant levels of anxiety symptoms. Although evidence suggests that anxious forms of depression are prognostic of poorer antidepressant outcomes, there is no clear definition of anxious depression, and inferences about clinical outcomes are thus limited. Our objective was to compare and evaluate definitions of anxious depression and anxiety-related scales according to clinical and antidepressant outcome criteria. Method: A total of 1008 adults with a current diagnosis of single-episode or recurrent, nonpsychotic, major depressive disorder were assessed at baseline on clinical features. Participants were then randomised to one of three antidepressants and reassessed at 8 weeks regarding remission and response of the 17-item Hamilton Rating Scale Depression (HRSD17) and the 16-item Quick Inventory of Depressive Symptomatology (QIDS-SR16). Anxious depression was defined as major depressive disorder with one or more anxiety disorders or major depressive disorder with a HRSD17 anxiety/somatisation factor score ⩾7. Anxiety-related scales included the HRSD17 anxiety/somatisation factor and the 42-item Depression Anxiety Stress Scales (DASS42) anxiety and stress subscales. Results: Anxious depression definitions showed poor agreement (κ = 0.15) and the HRSD17 anxiety/somatisation factor was weakly correlated with both DASS42 anxiety ( r = 0.24) and stress subscales ( r = 0.20). Anxious depression definitions were also associated with few impairments on clinical features and did not predict poorer antidepressant treatment outcome. However, higher DASS42 anxiety predicted poorer HRSD17 and QIDS-SR16 remission, and item-level analysis found higher scores on items 9 (situational anxiety) and 23 (somatic anxiety) of the DASS42 predicted poorer treatment outcome, even after adjusting for covariates and multiple comparisons. Conclusion: Common definitions of anxious depression show poor agreement and do not predict poorer treatment outcome. Anxiety symptoms may be better characterised dimensionally using DASS42 when predicting treatment outcome.

Evian Gordon,1 A John Rush,2 Donna M Palmer,3,4 Taylor A Braund,3 William Rekshan1 1Brain Resource, San Francisco, CA, USA; 2Duke-NUS, Singapore; 3Brain Resource, Sydney, NSW, Australia; 4Brain Dynamics Center, Sydney Medical School – Westmead and Westmead Millennium Institute, The University of Sydney, Sydney, NSW, Australia Purpose: To evaluate the performance of a cognitive and emotional test battery in a representative sample of depressed outpatients to inform likelihood of remission over 8 weeks of treatment with each of three common antidepressant medications. Patients and methods: Outpatients 18–65 years old with nonpsychotic major depressive disorder (17 sites) were randomized to escitalopram, sertraline or venlafaxine-XR (extended release). Participants scored ≥12 on the baseline 16-item Quick Inventory of Depressive Symptomatology – Self-Report and completed 8 weeks of treatment. The baseline test battery measured cognitive and emotional status. Exploratory multivariate logistic regression models predicting remission (16-item Quick Inventory of Depressive Symptomatology – Self-Report score ≤5 at 8 weeks) were developed independently for each medication in subgroups stratified by age, sex, or cognitive and emotional test performance. The model with the highest cross-validated accuracy determined the participant proportion in each arm for whom remission could be predicted with an accuracy ≥10% above chance. The proportion for whom a prediction could be made with very high certainty (positive predictive value and negative predictive value exceeding 80%) was calculated by incrementally increasing test battery thresholds to predict remission/non-remission. Results: The test battery, individually developed for each medication, improved identification of remitting and non-remitting participants by ≥10% beyond chance for 243 of 467 participants. The overall remission rates were escitalopram: 40.8%, sertraline: 30.3%, and venlafaxine-XR: 31.1%. Within this subset for whom prediction exceeded chance, test battery thresholds established a negative predictive value of ≥80%, which identified 40.9% of participants not remitting on escitalopram, 77.1% of participants not remitting on sertraline, and 38.7% of participants not remitting on venlafaxine-XR (all including 20% false negatives). Conclusion: The test battery identified about 50% of each medication group as being ≥10% more or less likely to remit than by chance, and identified about 38% of individuals who did not remit with ≥80% certainty. Clinicians might choose to avoid this specific medication in these particular patients. Keywords: depression, treatment selection, cognitive tests, biomarkers, treatment prediction, antidepressant medication