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.
Dementia prevention interventions that address modifiable risk factors for dementia require extensive lifestyle and behavior changes. Strategies are needed to enhance engagement and personalization ...of the experience at a population level. Precision Population Brain Health aims to improve brain health across the lifespan at a population level. Psychographic segmentation is a core component of Precision Population Brain Health with untapped potential. Psychographic segmentation applies behavioral and social sciences to understanding people's motivations, values, priorities, decision making, lifestyles, personalities, communication preferences, attitudes, and beliefs. Integrating psychographic segmentation into dementia care could provide a more personalized care experience and increased patient engagement, leading to improved health outcomes and reduced costs. Psychographic segmentation can enhance patient engagement for dementia and shift the clinical paradigm from "What is the matter?" to "What matters to you?" Similar benefits of psychographic segmentation can be provided for dementia caregivers. Developing dementia prevention programs that integrate psychographic segmentation could become the basis for creating a shared framework for prevention of non-communicable diseases and brain health disorders at a population level. Integrating psychographic segmentation into digital health tools for dementia prevention programs is especially critical to overcome current suboptimal approaches. Applying psychographic segmentation to dementia prevention has the potential to help people feel a sense of empowerment over their health and improve satisfaction with their health experience-creating a culture shift in the way brain health is approached and paving the way toward Precision Population Brain Health.
Depressed older adults are at risk for the development of mild cognitive impairment (MCI), but few studies have characterized MCI subtypes in geriatric depression. The objective of this study was to ...identify the clinical patterns of MCI in late-life depression.
Baseline demographic, clinical, and neuropsychological test data collected as part of a randomized antidepressant trial for geriatric depression.
UCLA-based outpatient clinic.
One hundred thirty-eight older adults with major depression.
A neuropsychological test battery and comprehensive evaluations of depression, apathy, quality of life, medical burden, and vascular risk factors.
Seventy-one participants (51%) had MCI and 67 (49%) were cognitively normal. Of subjects with MCI, 14 (20%) had amnestic MCI and 57 (80%) had non-amnestic MCI. Overall, patients with MCI had greater depression severity, poorer quality of life, and worse performance on the Mini-Mental State Exam than patients without MCI. Patients with non-amnestic MCI had significantly greater depression severity than patients without MCI. Across all subjects, depression severity correlated with impaired performance in language and visuospatial functioning.
Our findings suggest that MCI is associated with greater severity of depression, poorer quality of life, and worse global cognitive function. Overall, subtypes of MCI in geriatric depression differ in the patterns of functional impairment, which may require different therapeutic approaches.
Exercise and diet impact body composition, but their age-related brain effects are unclear at the molecular imaging level. To address these issues, the authors determined whether body mass index ...(BMI), physical activity, and diet relate to brain positron emission tomography (PET) of amyloid plaques and tau tangles using 2-(1-(6-(2-F-18fluoroethyl)(methyl)amino-2-naphthyl)ethylidene)malononitrile (FDDNP).
Volunteers (N = 44; mean age: 62.6 ± 10.7 years) with subjective memory impairment (N = 24) or mild cognitive impairment (MCI; N = 20) were recruited by soliciting for memory complaints. Levels of physical activity and extent of following a Mediterranean-type diet were self-reported. FDDNP-PET scans assessed plaque/tangle binding in Alzheimer disease-associated regions (frontal, parietal, medial and lateral temporal, posterior cingulate). Mixed models controlling for known covariates examined BMI, physical activity, and diet in relation to FDDNP-PET.
MCI subjects with above normal BMI (>25) had higher FDDNP-PET binding compared with those with normal BMI (1.11(0.03) versus 1.08(0.03), ES = 1.04, t(35) = 3.3, p = 0.002). Greater physical activity was associated with lower FDDNP-PET binding in MCI subjects (1.07(0.03) versus 1.11(0.03), ES = 1.13, t(35) = -3.1, p = 0.004) but not in subjects with subjective memory impairment (1.07(0.03) versus 1.07(0.03), ES = 0.02, t(35) = -0.1, p = 0.9). Healthier diet related to lower FDDNP-PET binding, regardless of cognitive status (1.07(0.03) versus 1.09(0.02), ES = 0.72, t(35) = -2.1, p = 0.04).
These preliminary findings are consistent with a relationship between risk modifiersand brain plaque/tangle deposition in nondemented individuals and supports maintenance of normal body weight, regular physical activity, and healthy diet to protect the brain during aging. (clinicaltrials.gov; NCT00355498).
This study examined the potential of an antidepressant drug, escitalopram, to improve depression, resilience to stress, and quality of life in family dementia caregivers in a randomized ...placebo-controlled double-blinded trial.
Forty family caregivers (43-91 years of age, 25 children and 15 spouses; 26 women) who were taking care of their relatives with Alzheimer disease were randomized to receive either escitalopram 10 mg/day or placebo for 12 weeks. Severity of depression, resilience, burden, distress, quality of life, and severity of care-recipient's cognitive and behavioral disturbances were assessed at baseline and over the course of the study. The Hamilton Depression Rating Scale scores at baseline ranged between 10 and 28. The groups were stratified by the diagnosis of major and minor depression.
Most outcomes favored escitalopram over placebo. The severity of depression improved, and the remission rate was greater with the drug compared with placebo. Measures of anxiety, resilience, burden, and distress improved on escitalopram compared with placebo.
Among caregivers, this small randomized controlled trial found that escitalopram use resulted in improvement in depression, resilience, burden and distress, and quality of life. Our results need to be confirmed in a larger sample.
Objectives
Default mode network (DMN) connectivity is altered in depression. We evaluated the relationship between changes in within‐network DMN connectivity and improvement in depression in a ...subsample of our parent clinical trial comparing escitalopram/memantine (ESC/MEM) to escitalopram/placebo (ESC/PBO) in older depressed adults (NCT01902004).
Methods
Twenty‐six participants with major depression (age > 60 years) and subjective memory complaints underwent treatment with ESC/MEM (n = 13) or ESC/PBO (n = 13), and completed baseline and 3‐month follow‐up resting state magnetic resonance imaging scans. Multi‐block partial least squares correlation analysis was used to evaluate the impact of treatment on within‐network DMN connectivity changes and their relationship with symptom improvement at 3 months (controlling for age and sex).
Results
A significant latent variable was identified, reflecting within‐network DMN connectivity changes correlated with symptom improvement (p = .01). Specifically, although overall group differences in within‐network DMN connectivity changes failed to reach significance, increased within‐network connectivity of posterior/lateral DMN regions (precuneus, angular gyrus, superior/middle temporal cortex) was more strongly and positively correlated with symptom improvement in the ESC/MEM group (r = 0.97, 95% confidence interval: 0.86–0.98) than in the ESC/PBO group (r = 0.36, 95% confidence interval: 0.13–0.72).
Conclusions
Increased within‐network connectivity of core DMN nodes was more strongly correlated with depressive symptom improvement with ESC/MEM than with ESC/PBO, supporting an improved engagement of brain circuitry implicated in the amelioration of depressive symptoms with combined ESC/MEM treatment in older adults with depression and subjective memory complaints.
We evaluated the relationship between changes in DMN connectivity and improvement in depression in a subsample of our parent clinical trial comparing escitalopram/memantine (ESC/MEM) to escitalopram/placebo (ESC/PBO) in older depressed adults (NCT01902004). A significant latent variable was identified, reflecting DMN connectivity changes related to symptom improvement. Increased connectivity of core DMN nodes was more strongly associated with depressive symptom improvement with ESC/MEM than with ESC/PBO, supporting a neuroplastic effect of memantine addition to escitalopram on brain circuitry that can help reduce depressive symptoms and prevent cognitive decline.