Lucija Abramovic

Can structural MRI aid in clinical classification? A machine learning study in two independent samples of patients with schizophrenia, bipolar disorder and healthy subjects.

Although structural magnetic resonance imaging (MRI) has revealed partly non-overlapping brain abnormalities in schizophrenia and bipolar disorder, it is unknown whether structural MRI scans can be used to separate individuals with schizophrenia from those with bipolar disorder. An algorithm capable of discriminating between these two disorders could become a diagnostic aid for psychiatrists. Here, we scanned 66 schizophrenia patients, 66 patients with bipolar disorder and 66 healthy subjects on a 1.5T MRI scanner. Three support vector machines were trained to separate patients with schizophrenia from healthy subjects, patients with schizophrenia from those with bipolar disorder, and patients with bipolar disorder from healthy subjects, respectively, based on their gray matter density images. The predictive power of the models was tested using cross-validation and in an independent validation set of 46 schizophrenia patients, 47 patients with bipolar disorder and 43 healthy subjects scanned on a 3T MRI scanner. Schizophrenia patients could be separated from healthy subjects with an average accuracy of 90%. Additionally, schizophrenia patients and patients with bipolar disorder could be distinguished with an average accuracy of 88%.The model delineating bipolar patients from healthy subjects was less accurate, correctly classifying 67% of the healthy subjects and only 53% of the patients with bipolar disorder. In the latter group, lithium and antipsychotics use had no influence on the classification results. Application of the 1.5T models on the 3T validation set yielded average classification accuracies of 76% (healthy vs schizophrenia), 66% (bipolar vs schizophrenia) and 61% (healthy vs bipolar). In conclusion, the accurate separation of schizophrenia from bipolar patients on the basis of structural MRI scans, as demonstrated here, could be of added value in the differential diagnosis of these two disorders. The results also suggest that gray matter pathology in schizophrenia and bipolar disorder differs to such an extent that they can be reliably differentiated using machine learning paradigms.

Subcortical volumetric abnormalities in bipolar disorder.

Considerable uncertainty exists about the defining brain changes associated with bipolar disorder (BD). Understanding and quantifying the sources of uncertainty can help generate novel clinical hypotheses about etiology and assist in the development of biomarkers for indexing disease progression and prognosis. Here we were interested in quantifying case–control differences in intracranial volume (ICV) and each of eight subcortical brain measures: nucleus accumbens, amygdala, caudate, hippocampus, globus pallidus, putamen, thalamus, lateral ventricles. In a large study of 1710 BD patients and 2594 healthy controls, we found consistent volumetric reductions in BD patients for mean hippocampus (Cohen’s d=−0.232; P=3.50 × 10−7) and thalamus (d=−0.148; P=4.27 × 10−3) and enlarged lateral ventricles (d=−0.260; P=3.93 × 10−5) in patients. No significant effect of age at illness onset was detected. Stratifying patients based on clinical subtype (BD type I or type II) revealed that BDI patients had significantly larger lateral ventricles and smaller hippocampus and amygdala than controls. However, when comparing BDI and BDII patients directly, we did not detect any significant differences in brain volume. This likely represents similar etiology between BD subtype classifications. Exploratory analyses revealed significantly larger thalamic volumes in patients taking lithium compared with patients not taking lithium. We detected no significant differences between BDII patients and controls in the largest such comparison to date. Findings in this study should be interpreted with caution and with careful consideration of the limitations inherent to meta-analyzed neuroimaging comparisons.

Increased psychophysiological parameters of attention in non-psychotic individuals with auditory verbal hallucinations

OBJECTIVE Schizophrenia is associated with aberrant event-related potentials (ERPs) such as reductions in P300, processing negativity and mismatch negativity amplitudes. These deficits may be related to the propensity of schizophrenia patients to experience auditory verbal hallucinations (AVH). However, AVH are part of extensive and variable symptomatology in schizophrenia. For this reason non-psychotic individuals with AVH as an isolated symptom provide an excellent opportunity to investigate this relationship. METHODS P300 waveforms, processing negativity and mismatch negativity were examined with an auditory oddball paradigm in 18 non-psychotic individuals with AVH and 18 controls. RESULTS P300 amplitude was increased in the AVH group as compared to controls, reflecting superior effortful attention. A trend in the same direction was found for processing negativity. No significant differences were found for mismatch negativity. CONCLUSION Contrary to our expectations, non-psychotic individuals with AVH show increased rather than decreased psychophysiological measures of effortful attention compared to healthy controls, refuting a pivotal role of decreased effortful attention in the pathophysiology of AVH.

High educational performance is a distinctive feature of bipolar disorder: a study on cognition in bipolar disorder, schizophrenia patients, relatives and controls

BACKGROUND Schizophrenia is associated with lower intelligence and poor educational performance relative to the general population. This is, to a lesser degree, also found in first-degree relatives of schizophrenia patients. It is unclear whether bipolar disorder I (BD-I) patients and their relatives have similar lower intellectual and educational performance as that observed in schizophrenia. METHOD This cross-sectional study investigated intelligence and educational performance in two outpatient samples [494 BD-I patients, 952 schizophrenia spectrum (SCZ) patients], 2231 relatives of BD-I and SCZ patients, 1104 healthy controls and 100 control siblings. Mixed-effects and regression models were used to compare groups on intelligence and educational performance. RESULTS BD-I patients were more likely to have completed the highest level of education (odds ratio 1.88, 95% confidence interval 1.66-2.70) despite having a lower IQ compared to controls (β = -9.09, S.E. = 1.27, p < 0.001). In contrast, SCZ patients showed both a lower IQ (β = -15.31, S.E. = 0.86, p < 0.001) and lower educational levels compared to controls. Siblings of both patient groups had significantly lower IQ than control siblings, but did not differ on educational performance. IQ scores did not differ between BD-I parents and SCZ parents, but BD-I parents had completed higher educational levels. CONCLUSIONS Although BD-I patients had a lower IQ than controls, they were more likely to have completed the highest level of education. This contrasts with SCZ patients, who showed both intellectual and educational deficits compared to healthy controls. Since relatives of BD-I patients did not demonstrate superior educational performance, our data suggest that high educational performance may be a distinctive feature of bipolar disorder patients.

Brain network analysis reveals affected connectome structure in bipolar I disorder

The notion that healthy brain function emerges from coordinated neural activity constrained by the brains network of anatomical connections—i.e., the connectome—suggests that alterations in the connectomes wiring pattern may underlie brain disorders. Corroborating this hypothesis, studies in schizophrenia are indicative of altered connectome architecture including reduced communication efficiency, disruptions of central brain hubs, and affected “rich club” organization. Whether similar deficits are present in bipolar disorder is currently unknown. This study examines structural connectome topology in 216 bipolar I disorder patients as compared to 144 healthy controls, focusing in particular on central regions (i.e., brain hubs) and connections (i.e., rich club connections, interhemispheric connections) of the brains network. We find that bipolar I disorder patients exhibit reduced global efficiency (−4.4%, P =0.002) and that this deficit relates (r = 0.56, P < 0.001) to reduced connectivity strength of interhemispheric connections (−13.0%, P = 0.001). Bipolar disorder patients were found not to show predominant alterations in the strength of brain hub connections in general, or of connections spanning brain hubs (i.e., “rich club” connections) in particular (all P > 0.1). These findings highlight a role for aberrant brain network architecture in bipolar I disorder with reduced global efficiency in association with disruptions in interhemispheric connectivity, while the central “rich club” system appears not to be particularly affected. Hum Brain Mapp 37:122–134, 2016.

Season of sampling and season of birth influence serotonin metabolite levels in human cerebrospinal fluid.

Background Animal studies have revealed seasonal patterns in cerebrospinal fluid (CSF) monoamine (MA) turnover. In humans, no study had systematically assessed seasonal patterns in CSF MA turnover in a large set of healthy adults. Methodology/Principal Findings Standardized amounts of CSF were prospectively collected from 223 healthy individuals undergoing spinal anesthesia for minor surgical procedures. The metabolites of serotonin (5-hydroxyindoleacetic acid, 5-HIAA), dopamine (homovanillic acid, HVA) and norepinephrine (3-methoxy-4-hydroxyphenylglycol, MPHG) were measured using high performance liquid chromatography (HPLC). Concentration measurements by sampling and birth dates were modeled using a non-linear quantile cosine function and locally weighted scatterplot smoothing (LOESS, span = 0.75). The cosine model showed a unimodal season of sampling 5-HIAA zenith in April and a nadir in October (p-value of the amplitude of the cosine = 0.00050), with predicted maximum (PCmax) and minimum (PCmin) concentrations of 173 and 108 nmol/L, respectively, implying a 60% increase from trough to peak. Season of birth showed a unimodal 5-HIAA zenith in May and a nadir in November (p = 0.00339; PCmax = 172 and PCmin = 126). The non-parametric LOESS showed a similar pattern to the cosine in both season of sampling and season of birth models, validating the cosine model. A final model including both sampling and birth months demonstrated that both sampling and birth seasons were independent predictors of 5-HIAA concentrations. Conclusion In subjects without mental illness, 5-HT turnover shows circannual variation by season of sampling as well as season of birth, with peaks in spring and troughs in fall.

The association of antipsychotic medication and lithium with brain measures in patients with bipolar disorder

There is evidence that brain structure is abnormal in patients with bipolar disorder. Lithium intake appears to ׳normalise׳ global and local brain volumes, but effects of antipsychotic medication on brain volume or cortical thickness are less clear. Here, we aim to disentangle disease-specific brain deviations from those induced by antipsychotic medication and lithium intake using a large homogeneous sample of patients with bipolar disorder type I. Magnetic resonance imaging brain scans were obtained from 266 patients and 171 control subjects. Subcortical volumes and global and focal cortical measures (volume, thickness, and surface area) were compared between patients and controls. In patients, the association between lithium and antipsychotic medication intake and global, subcortical and cortical measures was investigated. Patients showed significantly larger lateral and third ventricles, smaller total brain, caudate nucleus, and pallidum volumes and thinner cortex in some small clusters in frontal, parietal and cingulate regions as compared with controls. Lithium-free patients had significantly smaller total brain, thalamus, putamen, pallidum, hippocampus and accumbens volumes compared to patients on lithium. In patients, use of antipsychotic medication was related to larger third ventricle and smaller hippocampus and supramarginal cortex volume. Patients with bipolar disorder show abnormalities in total brain, subcortical, and ventricle volume, particularly in the nucleus caudate and pallidum. Abnormalities in cortical thickness were scattered and clusters were relatively small. Lithium-free patients showed more pronounced abnormalities as compared with those on lithium. The associations between antipsychotic medication and brain volume are subtle and less pronounced than those of lithium.

BDNF Val66Met homozygosity does not influence plasma BDNF levels in healthy human subjects

A putative pathway by which the BDNF Val66Met polymorphism (rs6265) leads to aberrant phenotypes is its influence on plasma BDNF. Research into the impact of rs6265 on plasma BDNF has given rise to conflicting results. Moreover, most such studies have compared Met-carriers with Val-homozygous subjects. We therefore genotyped subjects from a population-based cohort (the Utrecht Health Project, N=2743) and assessed whether plasma BDNF differs between rs6265 homozygous groups. We maximized the number of Met-homozygous subjects in whom we measured plasma BDNF, resulting in plasma BDNF being available for 19 Met-homozygous and 42 matched Val-homozygous subjects. Mean concentrations (S.D.) were 1963.1 (750.1) and 2133.2 pg/ml (1164.3) for the Val/Val and Met/Met groups, respectively. Using ANOVA, no differences in plasma BDNF between the two groups were detected. In conclusion, these results add to a growing body of evidence indicating that allelic variation at rs6265 does not have medium to large effects on plasma BDNF concentrations.

The relationship between brain volumes and intelligence in bipolar disorder

OBJECTIVES Bipolar disorder type-I (BD-I) patients show a lower Intelligence Quotient (IQ) and smaller brain volumes as compared with healthy controls. Considering that in healthy individuals lower IQ is related to smaller total brain volume, it is of interest to investigate whether IQ deficits in BD-I patients are related to smaller brain volumes and to what extent smaller brain volumes can explain differences between premorbid IQ estimates and IQ after a diagnosis of BD-I. METHODS Magnetic resonance imaging brain scans, IQ and premorbid IQ scores were obtained from 195 BDI patients and 160 controls. We studied the relationship of (global, cortical and subcortical) brain volumes with IQ and IQ change. Additionally, we investigated the relationship between childhood trauma, lithium- and antipsychotic use and IQ. RESULTS Total brain volume and IQ were positively correlated in the entire sample. This correlation did not differ between patients and controls. Although brain volumes mediated the relationship between BD-I and IQ in part, the direct relationship between the diagnosis and IQ remained significant. Childhood trauma and use of lithium and antipsychotic medication did not affect the relationship between brain volumes and IQ. However, current lithium use was related to lower IQ in patients. CONCLUSIONS Our data suggest a similar relationship between brain volume and IQ in BD-I patients and controls. Smaller brain volumes only partially explain IQ deficits in patients. Therefore, our findings indicate that in addition to brain volumes and lithium use other disease factors play a role in IQ deficits in BD-I patients.

The association of sleep and physical activity with integrity of white matter microstructure in bipolar disorder patients and healthy controls

We investigate how the sleep disruptions and irregular physical activity levels that are prominent features of bipolar disorder (BD) relate to white matter microstructure in patients and controls. Diffusion tension imaging (DTI) and 14-day actigraphy recordings were obtained in 51 BD I patients and 55 age-and-gender-matched healthy controls. Tract-based spatial statistics (TBSS) was used for voxelwise analysis of the association between fractional anisotropy (FA) and sleep and activity characteristics in the overall sample. Next, we investigated whether the relation between sleep and activity and DTI measures differed for patients and controls. Physical activity was related to increased integrity of white matter microstructure regardless of bipolar diagnosis. The relationship between sleep and white matter microstructure was more equivocal; we found an expected association between higher FA and effective sleep in controls but opposite patterns in bipolar patients. Confounding factors such as antipsychotic medication use are a likely explanation for these contrasting findings and highlight the need for further study of medication-related effects on white matter integrity.