Covariate adjusted classification trees.

In studies that compare several diagnostic groups, subjects can be measured on certain features and classification trees can be used to identify which of them best characterize the differences among groups. However, subjects may also be measured on additional covariates whose ability to characterize group differences is not meaningful or of interest, but may still have an impact on the examined features. Therefore, it is important to adjust for the effects of covariates on these features. We present a new semi-parametric approach to adjust for covariate effects when constructing classification trees based on the features of interest that is readily implementable. An application is given for postmortem brain tissue data to compare the neurobiological characteristics of subjects with schizophrenia to those of normal controls. We also evaluate the performance of our approach using a simulation study.

Functional Maturation of GABA Synapses During Postnatal Development of the Monkey Dorsolateral Prefrontal Cortex.

Development of inhibition onto pyramidal cells may be crucial for the emergence of cortical network activity, including gamma oscillations. In primate dorsolateral prefrontal cortex (DLPFC), inhibitory synaptogenesis starts in utero and inhibitory synapse density reaches adult levels before birth. However, in DLPFC, the expression levels of γ-aminobutyric acid (GABA) synapse-related gene products changes markedly during development until young adult age, suggesting a highly protracted maturation of GABA synapse function. Therefore, we examined the development of GABA synapses by recording GABAAR-mediated inhibitory postsynaptic currents (GABAAR-IPSCs) from pyramidal cells in the DLPFC of neonatal, prepubertal, peripubertal, and adult macaque monkeys. We found that the decay of GABAAR-IPSCs, possibly including those from parvalbumin-positive GABA neurons, shortened by prepubertal age, while their amplitude increased until the peripubertal period. Interestingly, both GABAAR-mediated quantal response size, estimated by miniature GABAAR-IPSCs, and the density of GABAAR synaptic appositions, measured with immunofluorescence microscopy, were stable with age. Simulations in a computational model network with constant GABA synapse density showed that the developmental changes in GABAAR-IPSC properties had a significant impact on oscillatory activity and predicted that, whereas DLPFC circuits can generate gamma frequency oscillations by prepubertal age, mature levels of gamma band power are attained at late stages of development.

Adjusting for matching and covariates in linear discriminant analysis.

In studies that compare several diagnostic or treatment groups, subjects may not only be measured on a certain set of feature variables, but also be matched on a number of demographic characteristics and measured on additional covariates. Linear discriminant analysis (LDA) is sometimes used to identify which feature variables best discriminate among groups, while accounting for the dependencies among the feature variables. We present a new approach to LDA for multivariate normal data that accounts for the subject matching used in a particular study design, as well as covariates not used in the matching. Applications are given for post-mortem tissue data with the aim of comparing neurobiological characteristics of subjects with schizophrenia with those of normal controls, and for a post-mortem tissue primate study comparing brain biomarker measurements across three treatment groups. We also investigate the performance of our approach using a simulation study.

A model for repeated clustered data with informative cluster sizes.

Many chronic diseases or health conditions manifest with recurring episodes, each of which can be characterized by a measure of intensity or severity. Both the number of episodes and the severity of each episode can depend on the latent severity of an individual's underlying condition. Data such as this are commonly gathered repeatedly at fixed follow-up intervals. An example is a study of the association between stressful life events and the onset of depression. Stress exposure is assessed through the frequency and intensity of stressful life events occurring each month. Both the number of events and the intensity of each event at each measurement occasion are informative about the underlying severity of stress over time. One might hypothesize that people that approach the onset of a depressive episode have worse stress profiles than the controls, reflected by both more frequent and more intense stressors. We propose models to analyze data collected repeatedly on both the frequency of an event and its severity when both of these are informative about the underlying latent severity. Maximum likelihood estimators are developed, and simulations with small to moderate sample sizes show that the estimators also have good finite sample properties, and they are robust against misspecification of the model. This method is applied to a psychiatric data set.

An adaptive two-stage dose-response design method for establishing proof of concept.

We propose an adaptive two-stage dose-response design where a prespecified adaptation rule is used to add and/or drop treatment arms between the stages. We extend the multiple comparison procedures-modeling (MCP-Mod) approach into a two-stage design. In each stage, we use the same set of candidate dose-response models and test for a dose-response relationship or proof of concept (PoC) via model-associated statistics. The stage-wise test results are then combined to establish "global" PoC using a conditional error function. Our simulation studies showed good and more robust power in our design method compared to conventional and fixed designs.

Increased expression of Kalirin-9 in the auditory cortex of schizophrenia subjects: its role in dendritic pathology.

Reductions in dendritic arbor length and complexity are among the most consistently replicated changes in neuronal structure in post mortem studies of cerebral cortical samples from subjects with schizophrenia, however, the underlying molecular mechanisms have not been identified. This study is the first to identify an alteration in a regulatory protein which is known to promote both dendritic length and arborization in developing neurons, Kalirin-9. We found Kalirin-9 expression to be paradoxically increased in schizophrenia. We followed up this observation by overexpressing Kalirin-9 in mature primary neuronal cultures, causing reduced dendritic length and complexity. Kalirin-9 overexpression represents a potential mechanism for dendritic changes seen in schizophrenia.

Simultaneous confidence bands for comparisons to placebo, with application to detecting the minimum effective dose.

In this article we consider a max-min approach to construct two-sided and one-sided simultaneous confidence bands in a dose-response study for the contrast in mean responses of each ascending dose versus placebo. The method utilizes the assumption of monotone non-decreasing dose-response curve. Also discussed is a step-down testing procedure that utilizes the lower bands to estimate the minimum effective dose (MED). The performance of our proposed step-down procedure is assessed by simulations and is shown to be superior to competitors when the MED is the lowest dose in a dose-response study.

Expected loss functions as additional measures to assess performance of multiple testing procedures for combination drug dose finding.

There are several measures that are commonly used to assess performance of a multiple testing procedure (MTP). These measures include power, overall error rate (family-wise error rate), and lack of power. In settings where the MTP is used to estimate a parameter, for example, the minimum effective dose, bias is of interest. In some studies, the parameter has a set-like structure, and thus, bias is not well defined. Nevertheless, the accuracy of estimation is one of the essential features of an MTP in such a context. In this paper, we propose several measures based on the expected values of loss functions that resemble bias. These measures are constructed to be useful in combination drug dose response studies when the target is to identify all minimum efficacious drug combinations. One of the proposed measures allows for assigning different penalties for incorrectly overestimating and underestimating a true minimum efficacious combination. Several simple examples are considered to illustrate the proposed loss functions. Then, the expected values of these loss functions are used in a simulation study to identify the best procedure among several methods used to select the minimum efficacious combinations, where the measures take into account the investigator's preferences about possibly overestimating and/or underestimating a true minimum efficacious combination. The ideas presented in this paper can be generalized to construct measures that resemble bias in other settings. These measures can serve as an essential tool to assess performance of several methods for identifying set-like parameters in terms of accuracy of estimation.

Chronic desipramine treatment alters tyrosine hydroxylase but not norepinephrine transporter immunoreactivity in norepinephrine axons in the rat prefrontal cortex.

Pharmacological blockade of norepinephrine (NE) reuptake is clinically effective in treating several mental disorders. Drugs that bind to the NE transporter (NET) alter both protein levels and activity of NET and also the catecholamine synthetic enzyme tyrosine hydroxylase (TH). We examined the rat prefrontal cortex (PFC) by electron microscopy to determine whether the density and subcellular distribution of immunolabelling for NET and co-localization of NET with TH within individual NE axons were altered by chronic treatment with the selective NE uptake inhibitor desipramine (DMI). Following DMI treatment (21 d, 15 mg/kg.d), NET-immunoreactive (ir) axons were significantly less likely to co-localize TH. This finding is consistent with reports of reduced TH levels and activity in the locus coeruleus after chronic DMI and indicates a reduction of NE synthetic capacity in the PFC. Measures of NET expression and membrane localization, including the number of NET-ir profiles per tissue area sampled, the number of gold particles per NET-ir profile area, and the proportion of gold particles associated with the plasma membrane, were similar in DMI- and vehicle-treated rats. These findings were verified using two different antibodies directed against distinct epitopes of the NET protein. The results suggest that chronic DMI treatment does not reduce NET expression within individual NE axons in vivo or induce an overall translocation of NET protein away from the plasma membrane in the PFC as measured by ultrastructural immunogold labelling. Our findings encourage consideration of possible post-translational mechanisms for regulating NET activity in antidepressant-induced modulation of NE clearance.

Simultaneous Confidence Band for the Difference of Segmented Linear Models.

Consider comparing between two treatments a response variable, whose expectation depends on the value of a continuous covariate in some nonlinear fashion. We fit separate segmented linear models to each treatment to approximate the nonlinear relationship. For this setting, we provide a simultaneous confidence band for the difference between treatments of the expected value functions. The treatments are said to differ significantly on intervals of the covariate where the simultaneous confidence band does not contain zero. We consider segmented linear models where the locations of the changepoints are both known and unknown. The band is obtained from asymptotic results.

Selective elimination of glutamatergic synapses on striatopallidal neurons in Parkinson disease models.

Parkinson disease is a common neurodegenerative disorder that leads to difficulty in effectively translating thought into action. Although it is known that dopaminergic neurons that innervate the striatum die in Parkinson disease, it is not clear how this loss leads to symptoms. Recent work has implicated striatopallidal medium spiny neurons (MSNs) in this process, but how and precisely why these neurons change is not clear. Using multiphoton imaging, we show that dopamine depletion leads to a rapid and profound loss of spines and glutamatergic synapses on striatopallidal MSNs but not on neighboring striatonigral MSNs. This loss of connectivity is triggered by a new mechanism-dysregulation of intraspine Cav1.3 L-type Ca(2+) channels. The disconnection of striatopallidal neurons from motor command structures is likely to be a key step in the emergence of pathological activity that is responsible for symptoms in Parkinson disease.

Recovery from impaired working memory performance during chronic Δ-9-tetrahydrocannabinol administration to adolescent rhesus monkeys.

BACKGROUND: The relationship between adolescent cannabis use and susceptibility to persistent cognitive impairments is poorly understood. AIMS: We examined the effects of repeated exposure to Δ-9-tetrahydrocannabinol (THC) on reinforcement-related learning and performance of spatial working memory (WM) tasks of varying difficulty in adolescent monkeys. METHODS: Seven pairs of male adolescent rhesus monkeys, matched for baseline cognitive performance, received vehicle or THC intravenously 5 days/week for 12 months. Performance on 4-item spatial WM trials was assessed throughout the 12-month study period. At the 6-month time point, more difficult novel and distractor 8-item spatial WM trials were added. Residual effects on performance were determined 23 or 71 h after THC or vehicle administration throughout the study. RESULTS/OUTCOMES: Relative to vehicle-exposed animals, repeated THC exposure was initially associated with significantly slower improvement in performance accuracy on 4-item spatial WM trials; however, this performance difference gradually diminished such that by month 12, accuracy did not significantly differ between vehicle and THC groups. Similarly, for the novel and distractor 8-item trials introduced at month 6, performance accuracy improved more slowly in the THC than in the vehicle group, despite comparable performance between groups on the 4-item task during this same period. CONCLUSIONS/INTERPRETATION: These findings suggest that compared to vehicle exposure, THC exposure during adolescence impairs the reinforcement-related learning process required for improved performance on spatial WM tasks, but this impairment might be overcome with continued training, even in the face of ongoing THC exposure.

Volume and neuron number of the lateral geniculate nucleus in schizophrenia and mood disorders.

Subjects with schizophrenia show deficits in visual perception that suggest changes predominantly in the magnocellular pathway and/or the dorsal visual stream important for visiospatial perception. We previously found a substantial 25% reduction in neuron number of the primary visual cortex (Brodmann's area 17, BA17) in postmortem tissue from subjects with schizophrenia. Also, many studies have found reduced volume and neuron number of the pulvinar--the large thalamic association nucleus involved in higher-order visual processing. Here, we investigate if the lateral geniculate nucleus (LGN), the visual relay nucleus of the thalamus, has structural changes in schizophrenia. We used stereological methods based on unbiased principles of sampling (Cavalieri's principle and the optical fractionator) to estimate the total volume and neuron number of the magno- and parovocellular parts of the left LGN in postmortem brains from nine subjects with schizophrenia, seven matched normal comparison subjects and 13 subjects with mood disorders. No significant schizophrenia-related structural differences in volume or neuron number of the left LGN or its major subregions were found, but we did observe a significantly increased total volume of the LGN, and of the parvocellular lamina and interlaminar regions, in the mood group. These findings do not support the hypothesis that subjects with schizophrenia have structural changes in the LGN. Therefore, our previous observation of a schizophrenia-related reduction of the primary visual cortex is probably not secondary to a reduction in the LGN.

Mixture modeling with applications in schizophrenia research.

Finite mixture modeling, together with the EM algorithm, have been widely used in clustering analysis. Under such methods, the unknown group membership is usually treated as missing data. When the "complete data" (log-)likelihood function does not have an explicit solution, the simplicity of the EM algorithm breaks down. Authors, including Rai and Matthews (1993), Lange (1995a) and Titterington (1984), developed modified algorithms therefore. As motivated by research in a large neurobiological project, we propose in this paper a new variant of such modifications and show that it is self-consistent. Moreover, simulations are conducted to demonstrate that the new variant converges faster than its predecessors.

Drop-the-Losers Design: Binomial Case.

Drop-the-losers designs were introduced for normal distributions as a method of combining phase II and III clinical trials together under a single protocol with the purpose of more rapidly evaluating drugs by eliminating as much as possible the delays that typically occur between the two phases of clinical development. In the design, the sponsor would administer k treatments along with a control in the first stage. During a brief interim period, efficacy data would be used to select the best treatment (with a rule to deal with ties) for further evaluation against the control in a second stage. At the end of the study, data from both stages would be used to draw inferences about the selected treatment relative to the control with adjustments made for selection in between the two stages. Because the inferences are model based, exact confidence intervals can be determined for the parameter of interest. In the present case, the parameter of concern is the probability of a beneficial response that is dichotomous in nature.

MAP2 immunoreactivity deficit is conserved across the cerebral cortex within individuals with schizophrenia.

Several postmortem studies have reported lower levels of immunoreactivity (IR) for microtubule-associated protein 2 (MAP2) in several cortical regions of individuals with schizophrenia (SZ). However, whether this effect is conserved across multiple brain areas within an individual with SZ or if it is regionally-specific remains unclear. We characterized patterns of MAP2-IR across three cortical regions at different levels of the rostral-caudal axis within individual subjects with and without SZ. MAP2-IR levels were measured in deep layer 3 of dorsolateral prefrontal cortex (DLPFC), lateral intraparietal cortex (LIP), and primary visual cortex (V1). Postmortem tissue containing each cortical region was derived from 20 pairs of SZ subjects and nonpsychiatric comparison (NPC) subjects matched perfectly for sex, and as closely as possible for age and postmortem interval. MAP2-IR was assessed by quantitative fluorescence microscopy. We observed significantly lower levels of MAP2-IR in SZ subjects relative to NPC subjects, without a significant region by diagnosis interaction. Logs of the within-pair ratios (SZ:NPC) of MAP2-IR were significantly correlated across the three regions. These findings demonstrate that MAP2-IR deficits in SZ are consistent across three neocortical regions within individual subjects. This pattern of MAP2-IR deficit has implications for therapeutic development and future investigations of MAP2 pathology in SZ.

Density of small dendritic spines and microtubule-associated-protein-2 immunoreactivity in the primary auditory cortex of subjects with schizophrenia.

Previously, we demonstrated that dendritic spine density (DSD) in deep layer 3 of the primary auditory cortex (A1) is lower, due to having fewer small spines, in subjects with schizophrenia (SZ) than non-psychiatric control (NPC) subjects. We also previously demonstrated that microtubule-associated-protein-2 immunoreactivity (MAP2-IR) in A1 deep layer 3 is lower, and positively correlated with DSD, in SZ subjects. Here, we first sought to confirm these findings in an independent cohort of 25 SZ-NPC subject pairs (cohort 1). We used immunohistochemistry and confocal microscopy to measure DSD and MAP2-IR in A1 deep layer 3. Consistent with previous studies, both DSD and MAP2-IR were lower in SZ subjects. We then tested the hypothesis that MAP2-IR mediates the effect of SZ on DSD in a cohort of 45 SZ-NPC subject pairs (combined cohort) that included all subjects from cohort 1 and two previously studied cohorts. Based on the distribution of MAP2-IR values in NPC subjects, we categorized each SZ subject as having either low MAP2-IR (SZ MAP2-IR(low)) or normal MAP2-IR (SZ MAP2-IR(normal)). Among SZ MAP-IR(low) subjects, mean DSD was significantly lower than in NPC subjects. However, mean DSD did not differ between SZ MAP2-IR(normal) and NPC subjects. Moreover, MAP2-IR statistically mediated small spine differences, with lower MAP2-IR values associated with fewer small spines. Our findings confirm that low density of small spines and low MAP2-IR are robust SZ phenotypes and suggest that MAP2-IR mediates the effect of SZ on DSD.

Conserved regional patterns of GABA-related transcript expression in the neocortex of subjects with schizophrenia.

OBJECTIVE: Individuals with schizophrenia exhibit disturbances in a number of cognitive, affective, sensory, and motor functions that depend on the circuitry of different cortical areas. The cognitive deficits associated with dysfunction of the dorsolateral prefrontal cortex result, at least in part, from abnormalities in GABA neurotransmission, as reflected in a specific pattern of altered expression of GABA-related genes. Consequently, the authors sought to determine whether this pattern of altered gene expression is restricted to the dorsolateral prefrontal cortex or could also contribute to the dysfunction of other cortical areas in subjects with schizophrenia. METHOD: Real-time quantitative polymerase chain reaction was used to assess the levels of eight GABA-related transcripts in four cortical areas (dorsolateral prefrontal cortex, anterior cingulate cortex, and primary motor and primary visual cortices) of subjects (N=12) with schizophrenia and matched normal comparison subjects. RESULTS: Expression levels of seven transcripts were lower in subjects with schizophrenia, with the magnitude of reduction for each transcript comparable across the four areas. The largest reductions were detected for mRNA encoding somatostatin and parvalbumin, followed by moderate decreases in mRNA expression for the 67-kilodalton isoform of glutamic acid decarboxylase, the GABA membrane transporter GAT-1, and the alpha 1 and delta subunits of GABA(A) receptors. In contrast, the expression of calretinin mRNA did not differ between the subject groups in any of the four areas. CONCLUSIONS: Because the areas examined represent the major functional domains (e.g., association, limbic, motor, and sensory) of the cerebral cortex, our findings suggest that a conserved set of molecular alterations affecting GABA neurotransmission contribute to the pathophysiology of different clinical features of schizophrenia.

Anatomical evidence of impaired feedforward auditory processing in schizophrenia.

BACKGROUND: Somal volumes of pyramidal cells are reduced within feedforward but not feedback circuits in areas 41 and 42 of the auditory cortex of subjects with schizophrenia. Because neuronal somal volume depends on both the number of axonal terminations onto and furnished by the neuron, we hypothesized that axon terminal densities are reduced in feedforward but not feedback auditory pathways in subjects with schizophrenia. METHODS: We used stereologic methods to quantify the density of a marker of axon terminals, synaptophysin-immunoreactive (SY-IR) puncta, in areas 41 and 42 of 15 subjects with schizophrenia and matched normal comparison subjects. The effect of long-term haloperidol exposure on density of SY-IR puncta was similarly evaluated in nonhuman primates. RESULTS: Synaptophysin-immunoreactive puncta density was 13.6% lower in deep layer 3 of area 41 in the schizophrenia subjects but was not changed in layer 1 of area 41 or in deep layer 3 of area 42. Density of SY-IR puncta did not differ between haloperidol-exposed and control monkeys. CONCLUSIONS: Reduction of SY-IR puncta density is selective for feedforward circuits within primary auditory cortex of subjects with schizophrenia. This deficit may contribute to impairments in auditory sensory processing in this disorder.

Primary visual cortex volume and total neuron number are reduced in schizophrenia.

A number of studies that assessed the visual system in subjects with schizophrenia found impairments in early visual processing. Furthermore, functional imaging studies suggested changes in primary visual cortex activity in subjects with schizophrenia. Interestingly, postmortem studies of subjects with schizophrenia reported an increased density of neurons in the primary visual cortex (Brodmann's area 17, BA17). The observed changes in visual processing may thus be reflected in structural changes in the circuitry of BA17. To characterize the structural changes further we used stereological methods based on unbiased principles of sampling (Cavalieri's principle and the optical fractionator) to estimate the total volume and neuron number of BA17 in postmortem brains from 10 subjects with schizophrenia and 10 matched normal comparison subjects. In addition, we assessed cortical thickness. We found a marked and significant reduction in total neuron number (25%) and volume (22%) of BA17 in the schizophrenia group relative to the normal comparison subjects. In contrast, we found no changes in neuronal density or cortical thickness between the two groups. Subjects with schizophrenia therefore have a smaller cortical area allocated to primary visual perception. This finding suggests the existence of a schizophrenia-related change in cortical parcellation.