Systematic evaluation of machine learning algorithms for neuroanatomically-based age prediction in youth.

Application of machine learning (ML) algorithms to structural magnetic resonance imaging (sMRI) data has yielded behaviorally meaningful estimates of the biological age of the brain (brain-age). The choice of the ML approach in estimating brain-age in youth is important because age-related brain changes in this age-group are dynamic. However, the comparative performance of the available ML algorithms has not been systematically appraised. To address this gap, the present study evaluated the accuracy (mean absolute error [MAE]) and computational efficiency of 21 machine learning algorithms using sMRI data from 2105 typically developing individuals aged 5-22 years from five cohorts. The trained models were then tested in two independent holdout datasets, one comprising 4078 individuals aged 9-10 years and another comprising 594 individuals aged 5-21 years. The algorithms encompassed parametric and nonparametric, Bayesian, linear and nonlinear, tree-based, and kernel-based models. Sensitivity analyses were performed for parcellation scheme, number of neuroimaging input features, number of cross-validation folds, number of extreme outliers, and sample size. Tree-based models and algorithms with a nonlinear kernel performed comparably well, with the latter being especially computationally efficient. Extreme Gradient Boosting (MAE of 1.49 years), Random Forest Regression (MAE of 1.58 years), and Support Vector Regression (SVR) with Radial Basis Function (RBF) Kernel (MAE of 1.64 years) emerged as the three most accurate models. Linear algorithms, with the exception of Elastic Net Regression, performed poorly. Findings of the present study could be used as a guide for optimizing methodology when quantifying brain-age in youth.

Sex differences in predictors and regional patterns of brain age gap estimates.

The brain-age-gap estimate (brainAGE) quantifies the difference between chronological age and age predicted by applying machine-learning models to neuroimaging data and is considered a biomarker of brain health. Understanding sex differences in brainAGE is a significant step toward precision medicine. Global and local brainAGE (G-brainAGE and L-brainAGE, respectively) were computed by applying machine learning algorithms to brain structural magnetic resonance imaging data from 1113 healthy young adults (54.45% females; age range: 22-37 years) participating in the Human Connectome Project. Sex differences were determined in G-brainAGE and L-brainAGE. Random forest regression was used to determine sex-specific associations between G-brainAGE and non-imaging measures pertaining to sociodemographic characteristics and mental, physical, and cognitive functions. L-brainAGE showed sex-specific differences; in females, compared to males, L-brainAGE was higher in the cerebellum and brainstem and lower in the prefrontal cortex and insula. Although sex differences in G-brainAGE were minimal, associations between G-brainAGE and non-imaging measures differed between sexes with the exception of poor sleep quality, which was common to both. While univariate relationships were small, the most important predictor of higher G-brainAGE was self-identification as non-white in males and systolic blood pressure in females. The results demonstrate the value of applying sex-specific analyses and machine learning methods to advance our understanding of sex-related differences in factors that influence the rate of brain aging and provide a foundation for targeted interventions.

Subcortical volumes across the lifespan: Data from 18,605 healthy individuals aged 3-90 years.

Age has a major effect on brain volume. However, the normative studies available are constrained by small sample sizes, restricted age coverage and significant methodological variability. These limitations introduce inconsistencies and may obscure or distort the lifespan trajectories of brain morphometry. In response, we capitalized on the resources of the Enhancing Neuroimaging Genetics through Meta-Analysis (ENIGMA) Consortium to examine age-related trajectories inferred from cross-sectional measures of the ventricles, the basal ganglia (caudate, putamen, pallidum, and nucleus accumbens), the thalamus, hippocampus and amygdala using magnetic resonance imaging data obtained from 18,605 individuals aged 3-90 years. All subcortical structure volumes were at their maximum value early in life. The volume of the basal ganglia showed a monotonic negative association with age thereafter; there was no significant association between age and the volumes of the thalamus, amygdala and the hippocampus (with some degree of decline in thalamus) until the sixth decade of life after which they also showed a steep negative association with age. The lateral ventricles showed continuous enlargement throughout the lifespan. Age was positively associated with inter-individual variability in the hippocampus and amygdala and the lateral ventricles. These results were robust to potential confounders and could be used to examine the functional significance of deviations from typical age-related morphometric patterns.

Cortical thickness across the lifespan: Data from 17,075 healthy individuals aged 3-90 years.

Delineating the association of age and cortical thickness in healthy individuals is critical given the association of cortical thickness with cognition and behavior. Previous research has shown that robust estimates of the association between age and brain morphometry require large-scale studies. In response, we used cross-sectional data from 17,075 individuals aged 3-90 years from the Enhancing Neuroimaging Genetics through Meta-Analysis (ENIGMA) Consortium to infer age-related changes in cortical thickness. We used fractional polynomial (FP) regression to quantify the association between age and cortical thickness, and we computed normalized growth centiles using the parametric Lambda, Mu, and Sigma method. Interindividual variability was estimated using meta-analysis and one-way analysis of variance. For most regions, their highest cortical thickness value was observed in childhood. Age and cortical thickness showed a negative association; the slope was steeper up to the third decade of life and more gradual thereafter; notable exceptions to this general pattern were entorhinal, temporopolar, and anterior cingulate cortices. Interindividual variability was largest in temporal and frontal regions across the lifespan. Age and its FP combinations explained up to 59% variance in cortical thickness. These results may form the basis of further investigation on normative deviation in cortical thickness and its significance for behavioral and cognitive outcomes.

Linked patterns of biological and environmental covariation with brain structure in adolescence: a population-based longitudinal study.

Adolescence is a period of major brain reorganization shaped by biologically timed and by environmental factors. We sought to discover linked patterns of covariation between brain structural development and a wide array of these factors by leveraging data from the IMAGEN study, a longitudinal population-based cohort of adolescents. Brain structural measures and a comprehensive array of non-imaging features (relating to demographic, anthropometric, and psychosocial characteristics) were available on 1476 IMAGEN participants aged 14 years and from a subsample reassessed at age 19 years (n = 714). We applied sparse canonical correlation analyses (sCCA) to the cross-sectional and longitudinal data to extract modes with maximum covariation between neuroimaging and non-imaging measures. Separate sCCAs for cortical thickness, cortical surface area and subcortical volumes confirmed that each imaging phenotype was correlated with non-imaging features (sCCA r range: 0.30-0.65, all PFDR < 0.001). Total intracranial volume and global measures of cortical thickness and surface area had the highest canonical cross-loadings (|ρ| = 0.31-0.61). Age, physical growth and sex had the highest association with adolescent brain structure (|ρ| = 0.24-0.62); at baseline, further significant positive associations were noted for cognitive measures while negative associations were observed at both time points for prenatal parental smoking, life events, and negative affect and substance use in youth (|ρ| = 0.10-0.23). Sex, physical growth and age are the dominant influences on adolescent brain development. We highlight the persistent negative influences of prenatal parental smoking and youth substance use as they are modifiable and of relevance for public health initiatives.

Impact of opioid agonist treatment on mental health in patients with opioid use disorder: a systematic review and network meta-analysis of randomized clinical trials.

Background: There is a knowledge gap in systematic reviews on the impact of opioid agonist treatments on mental health.Objectives: We compared mental health outcomes between different opioid agonist treatments and placebo/waitlist, and between the different opioids themselves.Methods: This meta-analysis of randomized clinical trials (RCTs) was pre-registered at PROSPERO (CRD42018109375). Embase, MEDLINE, PsychInfo, CINAHL Complete, and Web of Science Core Collection were searched from inception to May 2020. RCTs were included if they compared opioid agonists with each other or with placebo/waitlist in the treatment of patients with opioid use disorder and reported at least one mental health outcome after 1-month post-baseline. Studies with psychiatric care, adjunct psychotropic medications, or unbalanced psychosocial services were excluded. The primary outcome was overall mental health symptomatology, e.g. Symptom Checklist 90 total score, between opioids and placebo/waitlist. Random effects models were used for all the meta-analyses.Results: Nineteen studies were included in the narrative synthesis and 15 in the quantitative synthesis. Hydromorphone, diacetylmorphine (DAM), methadone, slow-release oral morphine, buprenorphine, and placebo/waitlist were among the included interventions. Based on the network meta-analysis for primary outcomes, buprenorphine (SMD (CI95%) = -0.61 (-1.20, -0.11)), DAM (-1.40 (-2.70, -0.23)), and methadone (-1.20 (-2.30, -0.11)) were superior to waitlist/placebo on overall mental health. Further direct pairwise meta-analysis indicated that overall mental health improved more in DAM compared to methadone (-0.23 (-0.34, -0.13)).Conclusions: Opioid agonist treatments used for the treatment of opioid use disorder improve mental health independent of psychosocial services.

Parsing out the role of dopamine D4 receptor gene (DRD4) on alcohol-related phenotypes: A meta-analysis and systematic review.

Genetics account for moderate variation of individual differences in developing alcohol use disorder (AUD), but it is unclear which genetic variations contribute to AUD risk. One candidate gene investigated due to its association with AUD is the dopamine D4 receptor gene (DRD4), which contains a 48-base pair variable number tandem repeat (VNTR) in exon 3 of its coding region. To date, no quantitative synthesis of the published literature on the effects of DRD4 VNTR variation on alcohol-related phenotypes has been conducted. MEDLINE, Embase, Web of Science, and PsycInfo were searched for studies that reported on alcohol craving, alcohol consumption, severity of AUD, and case-control (AUD versus no diagnosis of AUD) studies in DRD4L (seven repeats or more) carriers compared with DRD4S (six repeats or less) homozygotes. Random-effects meta-analysis was used for all analyses. A pooled sample size of 655 to 13,360 of 28 studies were included. Compared with DRD4S homozygotes, DRD4L carriers had increased number of drinking days (SMD: 0.205; 95% CI: 0.008 to 0.402), binge drinking days (SMD: 0.217; 95% CI: 0.0532 to 0.380), and severity of AUD (SMD: 0.143; 95% CI: 0.028 to 0.259). There was no difference between DRD4 VNTR genotypes on drinks per drinking day, largest number of drinks per day/occasion, and case-control analysis. It was not possible to conduct a meta-analysis of the craving data, but a systematic review of this literature found mixed results on DRD4 VNTR genotype effect. The present meta-analysis suggests DRD4 VNTR variation may be a risk factor for problematic alcohol use. Our findings are limited, however, by the absence of ancestry data from studies included in our analysis, precluding our ability to adjust for population stratification. Due to the likelihood of type I error in candidate gene approaches, our work highlights the critical need for studies with larger and more inclusive samples that account for sex and genetic ancestry to fully understand this relationship.

Apgar score and risk of autism.

Low Apgar score has been associated with higher risk for several neurological and psychiatric disorders, including cerebral palsy and intellectual disability. Studies of the association between Apgar score and autism spectrum disorder (ASD) have been inconsistent. We aimed to investigate (1) the association between low Apgar score at 5 min and risk for ASD, and (2) the modifying effects of gestational age and sex on this association in the largest multinational database of ASD. We included prospective data from 5.5 million individuals and over 33,000 cases of ASD from Norway, Sweden, Denmark and Western Australia who were born between 1984 and 2007. We calculated crude and adjusted risk ratios (RR) with 95% confidence intervals (95% CIs) for the associations between low Apgar score and ASD. All analyses for ASD were repeated for autistic disorder (AD). We used interaction terms and stratified analysis to investigate the effects of sex, gestational age, and birth weight on the association. In fully adjusted models, low Apgar scores (1-3) (RR, 1.42; 95% CI, 1.16-1.74), and intermediate Apgar scores (4-6) (RR, 1.50; 95% CI, 1.36-1.65) were associated with a higher RR of ASD than optimal Apgar score (7-10). The point estimates for low (RR, 1.88; 95% CI, 1.41-2.51) and intermediate Apgar score (RR, 1.54; 95% CI, 1.32-1.81) were larger for AD than for ASD. This study suggests that low Apgar score is associated with higher risk of ASD, and in particular AD. We did not observe any major modifying effects of gestational age and sex, although there seems to be substantial confounding by gestational age and birth weight on the observed association.

Environmental exposure to metals, neurodevelopment, and psychosis.

PURPOSE OF REVIEW: This article presents a new hypothesis about the possible relation between early life exposure to metals and psychosis. We review limitations of available research, and discuss novel approaches to overcome previous methodological barriers. RECENT FINDINGS: Mechanistic studies suggest a possible association between excess lead, manganese, cadmium, arsenic, or copper, and zinc deficiency, and several biochemical disturbances related to psychosis, such as altered neurotransmitters levels, excitotoxicity, and inflammation. Furthermore, studies suggest that some metals (lead, manganese, cadmium excess, and zinc deficiency) are associated with schizophrenia or psychosis-related phenotype. However, previous studies had multiple methodological limitations. Importantly, metal exposure was often measured after disease development and seldom determined during critical developmental periods. Most studies fell short of depicting the exact timing of exposure and the change in exposure over time. Here, we propose several methods to overcome these methodological limitations. SUMMARY: There is a plausible role of early life exposure to metals in the cause of psychosis. Owing to methodological limitations in exposure measurement, this has not been well characterized. Considering the wide exposure to metals and the high cost of psychosis to society, this hypothesis should be rigorously examined.

Memantine for Prophylactic Treatment of Migraine Without Aura: A Randomized Double-Blind Placebo-Controlled Study.

BACKGROUND: Uncontrolled studies in human have suggested that memantine might be a suitable option for migraine prophylaxis. OBJECTIVE: To assess the efficacy and tolerability of memantine for migraine prophylaxis. METHODS: This was a 12-week randomized double-blind placebo-controlled parallel-group study. Sixty patients with migraine without aura were randomized using a computer-generated list to receive memantine (10 mg/day) or placebo for 12 weeks. The primary outcome was the difference in change from baseline in the monthly attack frequency at week 12 between the two groups (using migraine diary). Secondary efficacy measures were assessed using several clinical, functional, and psychological instruments. We performed both complete case (CC) and intention-to-treat analyses (ITT). RESULTS: Twenty-five patients in the memantine group and 27 patients in the placebo group completed the study. Patients in the memantine group showed significantly greater reduction (mean change; 3.4; 95%CI, 2.3-4.4) in the monthly attack frequency than the placebo group (mean change, 1.0; 95%CI, 0.3-1.7) (mean difference [MD], 2.3; 95%CI, 1.1-3.5, P < .001). Both CC (MD, 4.9; 95%CI, 2.6-7.2 days), and ITT analyses (MD, 5.2; 95%CI, 2.0-8.5) showed significantly higher reduction in the mean number of migraine days in the memantine group than the placebo group (P < .01). Patients in the memantine group experienced greater reduction in the number of work absence days, severity, and disability score than the patients in the placebo group in both ITT and CC analyses. Changes in quality of life, sleep, depression, and anxiety did not differ between the two groups. Three patients in the memantine group complained of sedation, mild vertigo and nausea, and drowsiness. In the placebo group, one patient complained of nausea and another patient discontinued treatment after 2 weeks due to vertigo. CONCLUSION: Memantine might be a tolerable and efficacious option for prophylaxis in patients with migraine without aura. Tolerability, short duration required for titration, and safety profile in pregnancy might give memantine an advantage over other antimigraine medications. The study was registered in the Iranian Registry of Clinical Trials (Registration number: IRCT2013120115616N1).