Application of machine learning methods to describe the effects of conjugated equine estrogens therapy on region-specific brain volumes

Effects of adult male rat feminization treatments on brain morphology and metabolomic profile

Body feminization, as part of gender affirmation process of transgender women, decreases the volume of their cortical and subcortical brain structures. In this work, we implement a rat model of adult male feminization which reproduces the results in the human brain and allows for the longitudinal investigation of the underlying structural and metabolic determinants in the brain of adult male rats undergoing feminization treatments. Structural MRI and Diffusion Tensor Imaging (DTI) were used to non-invasively monitor in vivo cortical brain volume and white matter microstructure over 30 days in adult male rats receiving estradiol (E₂), estradiol plus cyproterone acetate (CA), an androgen receptor blocker and antigonadotropic agent (E₂ + CA), or vehicle (control). Ex vivo cerebral metabolic profiles were assessed by ¹H High Resolution Magic Angle Spinning NMR (¹H HRMAS) at the end of the treatments in samples from brain regions dissected after focused microwave fixation (5 kW). We found that; a) Groups receiving E₂ and E₂ + CA showed a generalized bilateral decrease in cortical volume; b) the E₂ + CA and, to a lesser extent, the E₂ groups maintained fractional anisotropy values over the experiment while these values decreased in the control group; c) E₂ treatment produced increases in the relative concentration of brain metabolites, including glutamate and glutamine and d) the glutamine relative concentration and fractional anisotropy were negatively correlated with total cortical volume. These results reveal, for the first time to our knowledge, that the volumetric decreases observed in trans women under cross-sex hormone treatment can be reproduced in a rat model. Estrogens are more potent drivers of brain changes in male rats than anti-androgen treatment.

The Link between Estradiol and Neuroplasticity in Transgender Women after Gender-Affirming Surgery: A Bimodal Hypothesis

For transgender individuals, gender-affirming surgery (GAS) and cross-sex hormone therapy (CSHT) are part of the gender transition process. Scientific evidence supporting the maintenance of CSHT after GAS-related gonadectomy is accumulating. However, few data are available on the impact of CSHT on the brain structure following hypogonadism. Thus, we aimed to investigate links between estradiol and brain cortical thickness (CTh) and cognition in 18 post-gonadectomy transgender women using a longitudinal design. For this purpose, the participants underwent a voluntary period of CSHT washout of at least 30 days, followed by estradiol re-institution for 60 days. High-resolution T1-weighted brain images, hormonal measures, working and verbal memory were collected at 2 time points: on the last day of the washout (t1) and on the last day of the 2-month CSHT period (t2). Between these 2 time points, CTh increased within the left precentral gyrus and right precuneus but decreased within the right lateral occipital cortex. However, these findings did not survive corrections of multiple comparisons. Nevertheless, there was a significant negative correlation between changes in estradiol levels and changes in CTh. This effect was evident in the left superior frontal gyrus, the left middle temporal gyrus, the right precuneus, the right superior temporal gyrus, and the right pars opercularis. Although there was an improvement in verbal memory following hypogonadism correction, we did not observe a significant relationship between changes in memory scores and CTh. Altogether, these findings suggest that there is a link between estradiol and CTh.

A novel random forest integrative approach based on endogenous CYP3A4 phenotype for predicting tacrolimus concentrations and dosages in Chinese renal transplant patients

WHAT IS KNOWN AND OBJECTIVE

Personalized treatment with tacrolimus has remained a challenge. The present study aimed to evaluate the potential of an integrative approach to predict individual tacrolimus concentrations and dosages based on endogenous CYP3A4 phenotype, CYP3A5 genotype and clinical variables.

METHODS

A random forest (RF) algorithm which incorporated an endogenous CYP3A4 phenotype (assessed by urinary ratio of 6β-hydroxycortisol and 6β-hydroxycortisone to cortisol and cortisone), CYP3A5*3 genotype and other clinical determinants of tacrolimus disposition was performed in 182 medically stable renal transplant recipients.

RESULTS AND DISCUSSION

The results suggested that endogenous CYP3A4 phenotype was the most important determinant of tacrolimus concentrations and dose requirements. RF models provided high goodness of fit (R² ) with .92 and .95 for the prediction of tacrolimus trough concentrations and dosages, respectively, as well as high predictability (Q² ) with 0.63 and 0.70, respectively. Significant correlations existed between experimental and predictive data.

WHAT IS NEW AND CONCLUSION

In summary, endogenous CYP3A4 phenotype is a critical biomarker for the determination of tacrolimus disposition. This predictive RF approach based on CYP3A4 biomarker with the combination of CYP3A5*3 genotype and other clinical variables can be used for predicting tacrolimus concentrations and dosages, which may serve as a useful tool in individualized tacrolimus dosing.

Neuroimaging studies in people with gender incongruence

The current review gives an overview of brain studies in transgender people. First, we describe studies into the aetiology of feelings of gender incongruence, primarily addressing the sexual differentiation hypothesis: does the brain of transgender individuals resemble that of their natal sex, or that of their experienced gender? Findings from neuroimaging studies focusing on brain structure suggest that the brain phenotypes of trans women (MtF) and trans men (FtM) differ in various ways from control men and women with feminine, masculine, demasculinized and defeminized features. The brain phenotypes of people with feelings of gender incongruence may help us to figure out whether sex differentiation of the brain is atypical in these individuals, and shed light on gender identity development. Task-related imaging studies may show whether brain activation and task performance in transgender people is sex-atypical. Second, we review studies that evaluate the effects of cross-sex hormone treatment on the brain. This type of research provides knowledge on how changes in sex hormone levels may affect brain structure and function.

Prediction of Incident Diabetes in the Jackson Heart Study Using High-Dimensional Machine Learning

Statistical models to predict incident diabetes are often based on limited variables. Here we pursued two main goals: 1) investigate the relative performance of a machine learning method such as Random Forests (RF) for detecting incident diabetes in a high-dimensional setting defined by a large set of observational data, and 2) uncover potential predictors of diabetes. The Jackson Heart Study collected data at baseline and in two follow-up visits from 5,301 African Americans. We excluded those with baseline diabetes and no follow-up, leaving 3,633 individuals for analyses. Over a mean 8-year follow-up, 584 participants developed diabetes. The full RF model evaluated 93 variables including demographic, anthropometric, blood biomarker, medical history, and echocardiogram data. We also used RF metrics of variable importance to rank variables according to their contribution to diabetes prediction. We implemented other models based on logistic regression and RF where features were preselected. The RF full model performance was similar (AUC = 0.82) to those more parsimonious models. The top-ranked variables according to RF included hemoglobin A1C, fasting plasma glucose, waist circumference, adiponectin, c-reactive protein, triglycerides, leptin, left ventricular mass, high-density lipoprotein cholesterol, and aldosterone. This work shows the potential of RF for incident diabetes prediction while dealing with high-dimensional data.

A Review of the Status of Brain Structure Research in Transsexualism

The present review focuses on the brain structure of male-to-female (MtF) and female-to-male (FtM) homosexual transsexuals before and after cross-sex hormone treatment as shown by in vivo neuroimaging techniques. Cortical thickness and diffusion tensor imaging studies suggest that the brain of MtFs presents complex mixtures of masculine, feminine, and demasculinized regions, while FtMs show feminine, masculine, and defeminized regions. Consequently, the specific brain phenotypes proposed for MtFs and FtMs differ from those of both heterosexual males and females. These phenotypes have theoretical implications for brain intersexuality, asymmetry, and body perception in transsexuals as well as for Blanchard's hypothesis on sexual orientation in homosexual MtFs. Falling within the aegis of the neurohormonal theory of sex differences, we hypothesize that cortical differences between homosexual MtFs and FtMs and male and female controls are due to differently timed cortical thinning in different regions for each group. Cross-sex hormone studies have reported marked effects of the treatment on MtF and FtM brains. Their results are used to discuss the early postmortem histological studies of the MtF brain.

Adult development and quality of life of transgender and gender nonconforming people

PURPOSE OF REVIEW

Research on the health of transgender and gender nonconforming people has been limited with most of the work focusing on transition-related care and HIV. The present review summarizes research to date on the overall development and quality of life of transgender and gender nonconforming adults, and makes recommendations for future research.

RECENT FINDINGS

Pervasive stigma and discrimination attached to gender nonconformity affect the health of transgender people across the lifespan, particularly when it comes to mental health and well-being. Despite the related challenges, transgender and gender nonconforming people may develop resilience over time. Social support and affirmation of gender identity play herein a critical role. Although there is a growing awareness of diversity in gender identity and expression among this population, a comprehensive understanding of biopsychosocial development beyond the gender binary and beyond transition is lacking.

SUMMARY

Greater visibility of transgender people in society has revealed the need to understand and promote their health and quality of life broadly, including but not limited to gender dysphoria and HIV. This means addressing their needs in context of their families and communities, sexual and reproductive health, and successful aging. Research is needed to better understand what factors are associated with resilience and how it can be effectively promoted.

Effects of Hormone Therapy on Brain Volumes Changes of Postmenopausal Women Revealed by Optimally-Discriminative Voxel-Based Morphometry

Backgrounds

The Women's Health Initiative Memory Study Magnetic Resonance Imaging (WHIMS-MRI) provides an opportunity to evaluate how menopausal hormone therapy (HT) affects the structure of older women’s brains. Our earlier work based on region of interest (ROI) analysis demonstrated potential structural changes underlying adverse effects of HT on cognition. However, the ROI-based analysis is limited in statistical power and precision, and cannot provide fine-grained mapping of whole-brain changes.

Methods

We aimed to identify local structural differences between HT and placebo groups from WHIMS-MRI in a whole-brain refined level, by using a novel method, named Optimally-Discriminative Voxel-Based Analysis (ODVBA). ODVBA is a recently proposed imaging pattern analysis approach for group comparisons utilizing a spatially adaptive analysis scheme to accurately locate areas of group differences, thereby providing superior sensitivity and specificity to detect the structural brain changes over conventional methods.

Results

Women assigned to HT treatments had significant Gray Matter (GM) losses compared to the placebo groups in the anterior cingulate and the adjacent medial frontal gyrus, and the orbitofrontal cortex, which persisted after multiple comparison corrections. There were no regions where HT was significantly associated with larger volumes compared to placebo, although a trend of marginal significance was found in the posterior cingulate cortical area. The CEE-Alone and CEE+MPA groups, although compared with different placebo controls, demonstrated similar effects according to the spatial patterns of structural changes.

Conclusions

HT had adverse effects on GM volumes and risk for cognitive impairment and dementia in older women. These findings advanced our understanding of the neurobiological underpinnings of HT effects.

Postmenopausal hormone therapy, type 2 diabetes mellitus, and brain volumes

OBJECTIVE

To examine whether the effect of postmenopausal hormone therapy (HT) on brain volumes in women aged 65-79 years differs depending on type 2 diabetes status during postintervention follow-up of a randomized controlled clinical trial.

METHODS

The Women's Health Initiative randomized clinical trials assigned women to HT (0.625 mg/day conjugated equine estrogens with or without 2.5 mg/day medroxyprogesterone acetate) or placebo for an average of 5.6 years. A total of 1,402 trial participants underwent brain MRI 2.4 years after the trials; these were repeated in 699 women 4.7 years later. General linear models were used to assess the interaction between diabetes status and HT assignment on brain volumes.

RESULTS

Women with diabetes at baseline or during follow-up who had been assigned to HT compared to placebo had mean decrement in total brain volume of -18.6 mL (95% confidence interval [CI] -29.6, -7.6). For women without diabetes, this mean decrement was -0.4 (95% CI -3.8, 3.0) (interaction p=0.002). This interaction was evident for total gray matter (p<0.001) and hippocampal (p=0.006) volumes. It was not evident for changes in brain volumes over follow-up or for ischemic lesion volumes and was not influenced by diabetes duration or oral medications.

CONCLUSIONS

For women aged 65 years or older who are at increased risk for brain atrophy due to type 2 diabetes, prescription of postmenopausal HT is associated with lower gray matter (total and hippocampal) volumes. Interactions with diabetes and insulin resistance may explain divergent findings on how estrogen influences brain volume among older women.

Morphological covariance in anatomical MRI scans can identify discrete neural pathways in the brain and their disturbances in persons with neuropsychiatric disorders

We hypothesize that coordinated functional activity within discrete neural circuits induces morphological organization and plasticity within those circuits. Identifying regions of morphological covariation that are independent of morphological covariation in other regions therefore may therefore allow us to identify discrete neural systems within the brain. Comparing the magnitude of these variations in individuals who have psychiatric disorders with the magnitude of variations in healthy controls may allow us to identify aberrant neural pathways in psychiatric illnesses. We measured surface morphological features by applying nonlinear, high-dimensional warping algorithms to manually defined brain regions. We transferred those measures onto the surface of a unit sphere via conformal mapping and then used spherical wavelets and their scaling coefficients to simplify the data structure representing these surface morphological features of each brain region. We used principal component analysis (PCA) to calculate covariation in these morphological measures, as represented by their scaling coefficients, across several brain regions. We then assessed whether brain subregions that covaried in morphology, as identified by large eigenvalues in the PCA, identified specific neural pathways of the brain. To do so, we spatially registered the subnuclei for each eigenvector into the coordinate space of a Diffusion Tensor Imaging dataset; we used these subnuclei as seed regions to track and compare fiber pathways with known fiber pathways identified in neuroanatomical atlases. We applied these procedures to anatomical MRI data in a cohort of 82 healthy participants (42 children, 18 males, age 10.5 ± 2.43 years; 40 adults, 22 males, age 32.42 ± 10.7 years) and 107 participants with Tourette's Syndrome (TS) (71 children, 59 males, age 11.19 ± 2.2 years; 36 adults, 21 males, age 37.34 ± 10.9 years). We evaluated the construct validity of the identified covariation in morphology using DTI data from a different set of 20 healthy adults (10 males, mean age 29.7 ± 7.7 years). The PCA identified portions of structures that covaried across the brain, the eigenvalues measuring the magnitude of the covariation in morphology along the respective eigenvectors. Our results showed that the eigenvectors, and the DTI fibers tracked from their associated brain regions, corresponded with known neural pathways in the brain. In addition, the eigenvectors that captured morphological covariation across regions, and the principal components along those eigenvectors, identified neural pathways with aberrant morphological features associated with TS. These findings suggest that covariations in brain morphology can identify aberrant neural pathways in specific neuropsychiatric disorders.

Hormone therapy for preventing cardiovascular disease in post-menopausal women

BACKGROUND

Evidence from systematic reviews of observational studies suggests that hormone therapy may have beneficial effects in reducing the incidence of cardiovascular disease events in post-menopausal women, however the results of randomised controlled trials (RCTs) have had mixed results. This is an updated version of a Cochrane review published in 2013.

OBJECTIVES

To assess the effects of hormone therapy for the prevention of cardiovascular disease in post-menopausal women, and whether there are differential effects between use in primary or secondary prevention. Secondary aims were to undertake exploratory analyses to (i) assess the impact of time since menopause that treatment was commenced (≥ 10 years versus < 10 years), and where these data were not available, use age of trial participants at baseline as a proxy (≥ 60 years of age versus < 60 years of age); and (ii) assess the effects of length of time on treatment.

SEARCH METHODS

We searched the following databases on 25 February 2014: Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library, MEDLINE, EMBASE and LILACS. We also searched research and trials registers, and conducted reference checking of relevant studies and related systematic reviews to identify additional studies.

SELECTION CRITERIA

RCTs of women comparing orally administered hormone therapy with placebo or a no treatment control, with a minimum of six months follow-up.

DATA COLLECTION AND ANALYSIS

Two authors independently assessed study quality and extracted data. We calculated risk ratios (RRs) with 95% confidence intervals (CIs) for each outcome. We combined results using random effects meta-analyses, and undertook further analyses to assess the effects of treatment as primary or secondary prevention, and whether treatment was commenced more than or less than 10 years after menopause.

MAIN RESULTS

We identified six new trials through this update. Therefore the review includes 19 trials with a total of 40,410 post-menopausal women. On the whole, study quality was good and generally at low risk of bias; the findings are dominated by the three largest trials. We found high quality evidence that hormone therapy in both primary and secondary prevention conferred no protective effects for all-cause mortality, cardiovascular death, non-fatal myocardial infarction, angina, or revascularisation. However, there was an increased risk of stroke in those in the hormone therapy arm for combined primary and secondary prevention (RR 1.24, 95% CI 1.10 to 1.41). Venous thromboembolic events were increased (RR 1.92, 95% CI 1.36 to 2.69), as were pulmonary emboli (RR 1.81, 95% CI 1.32 to 2.48) on hormone therapy relative to placebo.The absolute risk increase for stroke was 6 per 1000 women (number needed to treat for an additional harmful outcome (NNTH) = 165; mean length of follow-up: 4.21 years (range: 2.0 to 7.1)); for venous thromboembolism 8 per 1000 women (NNTH = 118; mean length of follow-up: 5.95 years (range: 1.0 to 7.1)); and for pulmonary embolism 4 per 1000 (NNTH = 242; mean length of follow-up: 3.13 years (range: 1.0 to 7.1)).We performed subgroup analyses according to when treatment was started in relation to the menopause. Those who started hormone therapy less than 10 years after the menopause had lower mortality (RR 0.70, 95% CI 0.52 to 0.95, moderate quality evidence) and coronary heart disease (composite of death from cardiovascular causes and non-fatal myocardial infarction) (RR 0.52, 95% CI 0.29 to 0.96; moderate quality evidence), though they were still at increased risk of venous thromboembolism (RR 1.74, 95% CI 1.11 to 2.73, high quality evidence) compared to placebo or no treatment. There was no strong evidence of effect on risk of stroke in this group. In those who started treatment more than 10 years after the menopause there was high quality evidence that it had little effect on death or coronary heart disease between groups but there was an increased risk of stroke (RR 1.21, 95% CI 1.06 to 1.38, high quality evidence) and venous thromboembolism (RR 1.96, 95% CI 1.37 to 2.80, high quality evidence).

AUTHORS' CONCLUSIONS

Our review findings provide strong evidence that treatment with hormone therapy in post-menopausal women overall, for either primary or secondary prevention of cardiovascular disease events has little if any benefit and causes an increase in the risk of stroke and venous thromboembolic events.

Application of random forests methods to diabetic retinopathy classification analyses

BACKGROUND

Diabetic retinopathy (DR) is one of the leading causes of blindness in the United States and world-wide. DR is a silent disease that may go unnoticed until it is too late for effective treatment. Therefore, early detection could improve the chances of therapeutic interventions that would alleviate its effects.

METHODOLOGY

Graded fundus photography and systemic data from 3443 ACCORD-Eye Study participants were used to estimate Random Forest (RF) and logistic regression classifiers. We studied the impact of sample size on classifier performance and the possibility of using RF generated class conditional probabilities as metrics describing DR risk. RF measures of variable importance are used to detect factors that affect classification performance.

PRINCIPAL FINDINGS

Both types of data were informative when discriminating participants with or without DR. RF based models produced much higher classification accuracy than those based on logistic regression. Combining both types of data did not increase accuracy but did increase statistical discrimination of healthy participants who subsequently did or did not have DR events during four years of follow-up. RF variable importance criteria revealed that microaneurysms counts in both eyes seemed to play the most important role in discrimination among the graded fundus variables, while the number of medicines and diabetes duration were the most relevant among the systemic variables.

CONCLUSIONS AND SIGNIFICANCE

We have introduced RF methods to DR classification analyses based on fundus photography data. In addition, we propose an approach to DR risk assessment based on metrics derived from graded fundus photography and systemic data. Our results suggest that RF methods could be a valuable tool to diagnose DR diagnosis and evaluate its progression.

Change in brain and lesion volumes after CEE therapies: the WHIMS-MRI studies

OBJECTIVES

To determine whether smaller brain volumes in older women who had completed Women's Health Initiative (WHI)-assigned conjugated equine estrogen-based hormone therapy (HT), reported by WHI Memory Study (WHIMS)-MRI, correspond to a continuing increased rate of atrophy an average of 6.1 to 7.7 years later in WHIMS-MRI2.

METHODS

A total of 1,230 WHI participants were contacted: 797 (64.8%) consented, and 729 (59%) were rescanned an average of 4.7 years after the initial MRI scan. Mean annual rates of change in total brain volume, the primary outcome, and rates of change in ischemic lesion volumes, the secondary outcome, were compared between treatment groups using mixed-effect models with adjustment for trial, clinical site, age, intracranial volumes, and time between MRI measures.

RESULTS

Total brain volume decreased an average of 3.22 cm(3)/y in the active arm and 3.07 cm(3)/y in the placebo arm (p = 0.53). Total ischemic lesion volumes increased in both arms at a rate of 0.12 cm(3)/y (p = 0.88).

CONCLUSIONS

Conjugated equine estrogen-based postmenopausal HT, previously assigned at WHI baseline, did not affect rates of decline in brain volumes or increases in brain lesion volumes during the 4.7 years between the initial and follow-up WHIMS-MRI studies. Smaller frontal lobe volumes were observed as persistent group differences among women assigned to active HT compared with placebo. Women with a history of cardiovascular disease treated with active HT, compared with placebo, had higher rates of accumulation in white matter lesion volume and total brain lesion volume. Further study may elucidate mechanisms that explain these findings.

Combining graph and machine learning methods to analyze differences in functional connectivity across sex

In this work we combine machine learning methods and graph theoretical analysis to investigate gender associated differences in resting state brain network connectivity. The set of all correlations computed from the fMRI resting state data is used as input features for classification. Two ensemble learning methods are used to perform the detection of the set of discriminative edges between groups (males vs. females) of brain networks: 1) Random Forest and 2) an ensemble method based on least angle shrinkage and selection operator (lasso) regressors. Permutation testing is used not only to assess significance of classification accuracy but also to evaluate significance of feature selection. Finally, these methods are applied to data downloaded from the Connectome Project website. Our results suggest that gender differences in brain function may be related to sexually dimorphic regional connectivity between specific critical nodes via gender-discriminative edges.

High dimensional classification of structural MRI Alzheimer's disease data based on large scale regularization

In this work we use a large scale regularization approach based on penalized logistic regression to automatically classify structural MRI images (sMRI) according to cognitive status. Its performance is illustrated using sMRI data from the Alzheimer Disease Neuroimaging Initiative (ADNI) clinical database. We downloaded sMRI data from 98 subjects (49 cognitive normal and 49 patients) matched by age and sex from the ADNI website. Images were segmented and normalized using SPM8 and ANTS software packages. Classification was performed using GLMNET library implementation of penalized logistic regression based on coordinate-wise descent optimization techniques. To avoid optimistic estimates classification accuracy, sensitivity, and specificity were determined based on a combination of three-way split of the data with nested 10-fold cross-validations. One of the main features of this approach is that classification is performed based on large scale regularization. The methodology presented here was highly accurate, sensitive, and specific when automatically classifying sMRI images of cognitive normal subjects and Alzheimer disease (AD) patients. Higher levels of accuracy, sensitivity, and specificity were achieved for gray matter (GM) volume maps (85.7, 82.9, and 90%, respectively) compared to white matter volume maps (81.1, 80.6, and 82.5%, respectively). We found that GM and white matter tissues carry useful information for discriminating patients from cognitive normal subjects using sMRI brain data. Although we have demonstrated the efficacy of this voxel-wise classification method in discriminating cognitive normal subjects from AD patients, in principle it could be applied to any clinical population.