Dataset Information

Deep learning-based automated speech detection as a marker of social functioning in late-life depression.

ABSTRACT:

Background

Late-life depression (LLD) is associated with poor social functioning. However, previous research uses bias-prone self-report scales to measure social functioning and a more objective measure is lacking. We tested a novel wearable device to measure speech that participants encounter as an indicator of social interaction.

Methods

Twenty nine participants with LLD and 29 age-matched controls wore a wrist-worn device continuously for seven days, which recorded their acoustic environment. Acoustic data were automatically analysed using deep learning models that had been developed and validated on an independent speech dataset. Total speech activity and the proportion of speech produced by the device wearer were both detected whilst maintaining participants' privacy. Participants underwent a neuropsychological test battery and clinical and self-report scales to measure severity of depression, general and social functioning.

Results

Compared to controls, participants with LLD showed poorer self-reported social and general functioning. Total speech activity was much lower for participants with LLD than controls, with no overlap between groups. The proportion of speech produced by the participants was smaller for LLD than controls. In LLD, both speech measures correlated with attention and psychomotor speed performance but not with depression severity or self-reported social functioning.

Conclusions

Using this device, LLD was associated with lower levels of speech than controls and speech activity was related to psychomotor retardation. We have demonstrated that speech activity measured by wearable technology differentiated LLD from controls with high precision and, in this study, provided an objective measure of an aspect of real-world social functioning in LLD.

SUBMITTER: Little B

PROVIDER: S-EPMC8311821 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Background: Recent evidence suggests that integration of multi-modal data improves performance in machine learning prediction of depression treatment outcomes. Here, we compared the predictive performance of three machine learning classifiers using differing combinations of sociodemographic characteristics, baseline clinical self-reports, cognitive tests, and structural magnetic resonance imaging (MRI) features to predict treatment outcomes in late-life depression (LLD). Methods: Data were combined from two clinical trials conducted with depressed adults aged 60 and older, including response to escitalopram (N = 32, NCT01902004) and Tai Chi (N = 35, NCT02460666). Remission was defined as a score of 6 or less on the 24-item Hamilton Rating Scale for Depression (HAMD) at the end of 24 weeks of treatment. Features subsets were constructed from baseline sociodemographic and clinical features, gray matter volumes (GMVs), or both. Three classification algorithms were compared: (1) Support Vector Machine-Radial Bias Function (SVMRBF), (2) Random Forest (RF), and (3) Logistic Regression (LR). A repeated 5-fold cross-validation approach with a wrapper-based feature selection method was used for model fitting. Model performance metrics included Area under the ROC Curve (AUC) and Matthews correlation coefficient (MCC). Cross-validated performance significance was tested by permutation analysis. Classifiers were compared by Cochran's Q and post-hoc pairwise comparisons using McNemar's Chi-Square test with Bonferroni correction. Results: For the RF and SVMRBF algorithms, the combined feature set outperformed the clinical and GMV feature sets with a final cross-validated AUC of 0.83 ± 0.11 and 0.80 ± 0.11, respectively. Both classifiers passed permutation analysis. The LR algorithm performed best using GMV features alone (AUC 0.79 ± 0.14) but failed to pass permutation analysis using any feature set. Performance of the three classifiers differed significantly for all three features sets. Important predictive features of treatment response included anterior and posterior cingulate volumes, depression characteristics, and self-reported health-related quality scores. Conclusion: This preliminary exploration into the use of ML and multi-modal data to identify predictors of general treatment response in LLD indicates that integration of clinical and structural MRI features significantly increases predictive capability. Identified features are among those previously implicated in geriatric depression, encouraging future work in this arena.

Project description:Late-life depression frequently co-occurs with cognitive impairment. To inform clinical management of these conditions, we examined the hypotheses that, relative to cognitively normal elders meeting DSM-IV criteria for major depressive disorder, those with cognitive impairment would require greater intensity of pharmacotherapy to reach criteria for antidepressant response and would take longer to respond.Using data from the MTLD-3 study, we conducted a series of secondary analyses examining the implications of cognitive impairment for short-term, open-trial pharmacotherapy of late-life depression (major depressive disorder in individuals 65 years and older). The treatment algorithm consisted of 3 steps: initial treatment with a selective serotonin reuptake inhibitor (SSRI), a switch to a serotonin-norepinephrine reuptake inhibitor (SNRI) if the patient did not respond, and addition of an atypical antipsychotic if the patient did not respond to the SNRI. The first subject entered the protocol in April 2004, and the last subject exited in September 2009. We examined data for participants who completed the acute phase of MTLD-3 as responders and received a cognitive diagnosis (N = 153) based on National Alzheimer's Coordinating Center (NACC) Uniform Data Set criteria. We divided participants into 3 groups on the basis of NACC cognitive diagnosis: no cognitive disorder (n = 74), mild cognitive impairment (n = 60), and dementia (n = 19). For each group, we calculated the proportion of participants requiring first- (SSRI), second- (SNRI), or third-step (add-on atypical antipsychotic) treatment to meet criteria for response (17-Item Hamilton Depression Rating Scale score ? 10 for 3 consecutive weeks). We compared time to response across groups and correlates of nonresponse.The 3 groups did not differ in intensity of pharmacotherapy (P = .68) or time to response (P = .84). Nonresponse was more strongly correlated with longer major depressive episode duration (P = .0015), presence of recurrent depression (P = .002), and younger current age (P = .047), rather than cognitive status (P = .61).Cognitive status does not appear to impact short-term pharmacotherapy response variability in individuals whose depression responds to treatment with open-trial antidepressants delivered in a supportive, university-based medication clinic.

Project description:BACKGROUND:Suicidal behavior is associated with impaired decision making in contexts of uncertainty. Existing studies, however, do not definitively address whether suidice attempers have 1) impairment in learning from experience or 2) impairment in choice based on comparison of estimated option values. Our reinforcement learning model-based behavioral study tested these hypotheses directly in middle-aged and older suicide attempters representative of those who die by suicide. METHODS:Two samples (sample 1, n = 135; sample 2, n = 125) of suicide attempters with depression (nattempters = 54 and 39, respectively), suicide ideators, nonsuicidal patients with depression, and healthy control participants completed a probabilistic three-choice decision-making task. A second experiment in sample 2 experimentally dissociated long-term learned value from reward magnitude. Analyses combined computational reinforcement learning and mixed-effects models of decision times and choices. RESULTS:With regard to learning, suicide attempters (vs. all comparison groups) were less sensitive to one-back reinforcement, as indicated by a reduced effect on both choices and decision times. Learning deficits scaled with attempt lethality and were partially explained by poor cognitive control. With regard to value-based choice, suicide attempters (vs. all comparison groups) displayed abnormally long decision times when choosing between similarly valued options and were less able to distinguish between the best and second-best options. Group differences in value-based choice were robust to controlling for cognitive performance, comorbidities, impulsivity, psychotropic exposure, and possible brain damage from attempts. CONCLUSIONS:Serious suicidal behavior is associated with impaired reward learning, likely undermining the search for alternative solutions. Attempted suicide is associated with impaired value comparison during the choice process, potentially interfering with the consideration of deterrents and alternatives in a crisis.

Project description:Late-life depression (LLD), compared to depression at a young age, is more likely to have poor prognosis and high risk of progression to dementia. A recent systemic review and meta-analysis of the present antidepressants for LLD showed that the treatment response rate was 48% and the remission rate was only 33.7%, thus implying the need to improve the treatment with other approaches in the future. Recently, agents modulating the glutamatergic system have been tested for mental disorders such as schizophrenia, dementia, and depressive disorder. Ketamine, a noncompetitive NMDA receptor (NMDAR) antagonist, requires more evidence from randomized clinical trials (RCTs) to prove its efficacy and safety in treating LLD. The metabotropic receptors (mGluRs) of the glutamatergic system are family G-protein-coupled receptors, and inhibition of the Group II mGluRs subtypes (mGlu2 and mGlu3) was found to be as effective as ketamine in exerting rapid antidepressant activity in some animal studies. Inflammation has been thought to contribute to depression for a long time. The cytokine levels not only increase with age but also decrease serotonin. Regarding LLD, interleukin 6 (IL-6) and tumor necrosis factor α (TNF-α) released in vivo are likely to contribute to the reduced serotonin level. Brain-derived neurotrophic factor (BDNF), a growth factor and a modulator in the tropomyosin receptor kinase (Trk) family of tyrosine kinase receptors, probably declines quantitatively with age. Recent studies suggest that BDNF/TrkB decrement may contribute to learning deficits and memory impairment. In the process of aging, physiological changes in combination with geriatric diseases such as vascular diseases result in poorer prognosis of LLD in comparison with that of young-age depression. Treatments with present antidepressants have been generally unsatisfactory. Novel treatments such as anti-inflammatory agents or NMDAR agonists/antagonists require more studies in LLD. Last but not least, LLD and dementia, which share common pathways and interrelate reciprocally, are a great concern. If it is possible to enhance the treatment of LDD, dementia can be prevented or delated.