David M. Nickerson

Subgroup Analysis via Recursive Partitioning

Subgroup analysis is an integral part of comparative analysis where assessing the treatment effect on a response is of central interest. Its goal is to determine the heterogeneity of the treatment effect across subpopulations. In this paper, we adapt the idea of recursive partitioning and introduce an interaction tree (IT) procedure to conduct subgroup analysis. The IT procedure automatically facilitates a number of objectively defined subgroups, in some of which the treatment effect is found prominent while in others the treatment has a negligible or even negative effect. The standard CART (Breiman et al., 1984) methodology is inherited to construct the tree structure. Also, in order to extract factors that contribute to the heterogeneity of the treatment effect, variable importance measure is made available via random forests of the interaction trees. Both simulated experiments and analysis of census wage data are presented for illustration.

Estimating Physiological Thresholds with Continuous Two-Phase Regression

Abrupt changes in the relationship between physiological responses and environmental parameters yield data that frequently cannot be described with a single regression equation. Many approaches used to deal with this problem result in incomplete description of the data and imprecise approximations of the physiological threshold(s) at which the relationship changes. We describe a technique for determining the best continuous two-phase, straight-line regression model and for statistically estimating the point at which the relationship between the independent and dependent variables changes (i.e., threshold point).

Expression of an Exogenous Human Oct-4 Promoter Identifies Tumor-Initiating Cells in Osteosarcoma

We explored the nature of the tumor-initiating cell in osteosarcoma, a bone malignancy that predominately occurs in children. Previously, we observed expression of Oct-4, an embryonal transcriptional regulator, in osteosarcoma cell cultures and tissues. To examine the relationship between Oct-4 and tumorigenesis, cells from an osteosarcoma biopsy (OS521) were stably transfected with a plasmid containing the human Oct-4 promoter driving a green fluorescent protein (GFP) reporter to generate the transgenic line OS521Oct-4p. In culture, only approximately 24% of the OS521Oct-4p cells were capable of activating the transgenic Oct-4 promoter; yet, xenograft tumors generated in NOD/SCID mice contained approximately 67% GFP(+) cells, which selectively expressed the mesenchymal stem cell-associated surface antigens CD105 and ICAM-1. Comparison of the tumor-forming capacity of GFP-enriched and GFP-depleted cell fractions revealed that the GFP-enriched fractions were at least 100-fold more tumorigenic, capable of forming tumors at doses of <300 cells, and formed metastases in the lung. Clonal populations derived from a single Oct-4/GFP(+) cell were capable of forming tumors heterogeneous for Oct-4/GFP expression. These data are consistent with the cancer stem cell model of tumorigenesis in osteosarcoma and implicate a functional link between the capacity to activate an exogenous Oct-4 promoter and tumor formation. This osteosarcoma tumor-initiating cell appears highly prolific and constitutes a majority of the cell population in a primary xenograft tumor, which may provide a biological basis for the particular virulence of this type of cancer.

Principal Component Analyses of Assemblage Structure Data: Utility of Tests Based on Eigenvalues

We examined the ability of eigenvalue tests to distinguish field-collected from random, assemblage structure data sets. Eight published time series of species abun- dances were used in the analysis, including data sets for: fishes, birds, mammals, stream benthos, and crabs. To test the efficacy of eigenvalue tests, we constructed 1000 randomly generated data sets for each real data set, whose means and variances were identical to the means and variances of the original data matrices. The data sets were then subjected to a principal components analysis (PCA) and eigenvalue tests used to identify significant ei- genvalues for both correlation and covariance matrix solutions. We also examined the effects of: (1) number of species (= number of variables), (2) number of samples (= rep- lication), and (3) variance structure, on the performance of the test. Using PCAs based on the correlation matrix and with sample sizes typically encountered in the field, the eigenvalue tests generally performed at the .05 level when a = .01. Slightly poorer results were obtained with the covariance matrix. Increasing the number of samples to at least three times the number of species generally gave a level coverage for an a level test (i.e., a = .05, .01). Increasing variance in the data set only affected test outcomes at levels of replication less than twice the number of species. We conclude that the eigenvalue tests can be used to detect patterns in PCAs of assemblage structure data, if the number of samples is at least three times the number of species and either a covariance or correlation matrix solution is used. It is assumed that these patterns represent ecologically meaningful patterns of variation.

Intra‐articular gene delivery and expression of interleukin‐1Ra mediated by self‐complementary adeno‐associated virus

The adeno‐associated virus (AAV) has many safety features that favor its use in the treatment of arthritic conditions; however, the conventional, single‐stranded vector is inefficient for gene delivery to fibroblastic cells that primarily populate articular tissues. This has been attributed to the inability of these cells to convert the vector to a double‐stranded form. To overcome this, we evaluated double‐stranded self‐complementary (sc) AAV as a vehicle for intra‐articular gene delivery.

GEOGRAPHICGRADIENTS IN DIET AFFECT POPULATION DYNAMICS OF CANADA LYNX

Geographical gradients in the stability of cyclic populations of herbivores and their predators may relate to the degree of specialization of predators. However, such changes are usually associated with transition from specialist to generalist predator species, rather than from geographical variation in dietary breadth of specialist predators. Canada lynx (Lynx canadensis) and snowshoe hare (Lepus americanus) populations undergo cyclic fluctuations in northern parts of their range, but cycles are either greatly attenuated or lost altogether in the southern boreal forest where prey diversity is higher. We tested the influence of prey specialization on population cycles by measuring the stable carbon and nitrogen isotope ratios in lynx and their prey, estimating the contribution of hares to lynx diet across their range, and correlating this degree of specialization to the strength of their population cycles. Hares dominated the lynx diet across their range, but specialization on hares decreased in southern and western populations. The degree of specialization correlated with cyclic signal strength indicated by spectral analysis of lynx harvest data, but overall variability of lynx harvest (the standard deviation of natural-log-transformed harvest numbers) did not change significantly with dietary specialization. Thus, as alternative prey became more important in the lynx diet, the fluctuations became decoupled from a regular cycle but did not become less variable. Our results support the hypothesis that alternative prey decrease population cycle regularity but emphasize that such changes may be driven by dietary shifts among dominant specialist predators rather than exclusively through changes in the predator community.

Climatic variability leads to later seasonal flowering of Floridian plants.

Understanding species responses to global change will help predict shifts in species distributions as well as aid in conservation. Changes in the timing of seasonal activities of organisms over time may be the most responsive and easily observable indicator of environmental changes associated with global climate change. It is unknown how global climate change will affect species distributions and developmental events in subtropical ecosystems or if climate change will differentially favor nonnative species. Contrary to previously observed trends for earlier flowering onset of plant species with increasing spring temperatures from mid and higher latitudes, we document a trend for delayed seasonal flowering among plants in Florida. Additionally, there were few differences in reproductive responses by native and nonnative species to climatic changes. We argue that plants in Florida have different reproductive cues than those from more northern climates. With global change, minimum temperatures have become more variable within the temperate-subtropical zone that occurs across the peninsula and this variation is strongly associated with delayed flowering among Florida plants. Our data suggest that climate change varies by region and season and is not a simple case of species responding to consistently increasing temperatures across the region. Research on climate change impacts need to be extended outside of the heavily studied higher latitudes to include subtropical and tropical systems in order to properly understand the complexity of regional and seasonal differences of climate change on species responses.

Construction of a Conservative Confidence Region from Projections of an Exact Confidence Region in Multiple Linear Regression

Abstract The problem of constructing a confidence region for simultaneously estimating p, p ≥ 2, linear regression parameters for which confidence statements can be made on the individual parameters is revisited. Here, an intercept may be included among the p paremeters. The technique is due to Working and Hotelling (1929) and Scheffe (1959) and uses the p separate projections of the exact (1 – α) 100% confidence ellipsoid (ellipse if p = 2) to give confidence intervals for each regression parameter. The Cartesian product of these p confidence intervals gives a p-dimensional rectangle that contains the confidence ellipsoid and hence has a joint confidence coefficient of at least (1–α). A simple calculus proof is given to determine these projections. The projection procedure is compared with the Bonferroni procedure for this case.

Premigration Persecution, Postmigration Stressors and Resources, and Postmigration Mental Health: A Study of Severely Traumatized U.S. Arab Immigrant Women

Background: Competing theories exist regarding the importance of premigration trauma as compared with postmigration stressors and resources with respect to the risk to immigrant mental health. Objective: To examine how type of premigration trauma, postmigration stressors, and postmigration resources differentially predict posttraumatic stress disorder (PTSD) and major depressive disorder (MDD) symptomatology in Arab immigrant women who have been exposed to premigration trauma. Design: Descriptive, using multinomial logistic regression to explain membership in one of four groups: (a) PTSD only (n = 14), (b) MDD (n = 162), (c) comorbid PTSD–MDD (n = 148), and (d) subclinical symptoms (n = 209). Results: Parameter estimates for postimmigration-related stressors (as measured by the Demands of Immigration [DI]) indicated that a unit increase in DI scores was associated with a nearly 16-fold increase in the likelihood of being in the comorbid relative to the subclinical group, and a nearly 2.5-fold increase in the likelihood of being in the comorbid relative to the MDD-only group (p < .05). Odds ratios for social support, age, and type of premigration trauma ranged between 0.95 and 1.95 and only differentiated between subclinical and comorbid PTSD–MDD groups (p < .05). Conclusion: Postmigration stressors exert substantive effects on immigrant mental health outcomes. Nursing interventions are needed to reduce immigration-related stressors. Screening Arab immigrant women for depression and PTSD is important, given the high levels observed in this community-based sample.

Perspectives on the use of gene therapy for chronic joint diseases.

Advances in molecular and cellular biology have identified a wide variety of proteins including targeted cytokine inhibitors, immunomodulatory proteins, cytotoxic mediators, angiogenesis inhibitors, and intracellular signalling molecules that could be of great benefit in the treatment of chronic joint diseases, such as osteo- and rheumatoid arthritis. Unfortunately, protein-based drugs are difficult to administer effectively. They have a high rate of turnover, requiring frequent readministration, and exposure in non-diseased tissue can lead to serious side effects. Gene transfer technologies offer methods to enhance the efficacy of protein-based therapies, enabling the body to produce these molecules locally at elevated levels for extended periods. The proof of concept of gene therapies for arthritis has been exhaustively demonstrated in multiple laboratories and in numerous animal models. This review attempts to condense these studies and to discuss the relative benefits and limitations of the methods proposed and to discuss the challenges toward translating these technologies into clinical realities.