Analyzing Multiple Outcomes: Is it Really Worth the use of Multivariate Linear Regression?

In health related research it is common to have multiple outcomes of interest in a single study. These outcomes are often analysed separately, ignoring the correlation between them. One would expect that a multivariate approach would be a more efficient alternative to individual analyses of each outcome. Surprisingly, this is not always the case. In this article we discuss different settings of linear models and compare the multivariate and univariate approaches. We show that for linear regression models, the estimates of the regression parameters associated with covariates that are shared across the outcomes are the same for the multivariate and univariate models while for outcome-specific covariates the multivariate model performs better in terms of efficiency.

MART-1 peptide vaccination plus IMP321 (LAG-3Ig fusion protein) in patients receiving autologous PBMCs after lymphodepletion: results of a Phase I trial.

BACKGROUND: Immunotherapy offers a promising novel approach for the treatment of cancer and both adoptive T-cell transfer and immune modulation lead to regression of advanced melanoma. However, the potential synergy between these two strategies remains unclear. METHODS: We investigated in 12 patients with advanced stage IV melanoma the effect of multiple MART-1 analog peptide vaccinations with (n = 6) or without (n = 6) IMP321 (LAG-3Ig fusion protein) as an adjuvant in combination with lymphodepleting chemotherapy and adoptive transfer of autologous PBMCs at day (D) 0 (Trial registration No: NCT00324623). All patients were selected on the basis of ex vivo detectable MART-1-specific CD8 T-cell responses and immunized at D0, 8, 15, 22, 28, 52, and 74 post-reinfusion. RESULTS: After immunization, a significant expansion of MART-1-specific CD8 T cells was measured in 83% (n = 5/6) and 17% (n = 1/6) of patients from the IMP321 and control groups, respectively (P < 0.02). Compared to the control group, the mean fold increase of MART-1-specific CD8 T cells in the IMP321 group was respectively >2-, >4- and >6-fold higher at D15, D30 and D60 (P < 0.02). Long-lasting MART-1-specific CD8 T-cell responses were significantly associated with IMP321 (P < 0.02). At the peak of the response, MART-1-specific CD8 T cells contained higher proportions of effector (CCR7⁻ CD45RA⁺/⁻) cells in the IMP321 group (P < 0.02) and showed no sign of exhaustion (i.e. were mostly PD1⁻CD160⁻TIM3⁻LAG3⁻2B4⁺/⁻). Moreover, IMP321 was associated with a significantly reduced expansion of regulatory T cells (P < 0.04); consistently, we observed a negative correlation between the relative expansion of MART-1-specific CD8 T cells and of regulatory T cells. Finally, although there were no confirmed responses as per RECIST criteria, a transient, 30-day partial response was observed in a patient from the IMP321 group. CONCLUSIONS: Vaccination with IMP321 as an adjuvant in combination with lymphodepleting chemotherapy and adoptive transfer of autologous PBMCs induced more robust and durable cellular antitumor immune responses, supporting further development of IMP321 as an adjuvant for future immunotherapeutic strategies.

Use of Two-Part Regression Calibration Model to Correct for Measurement Error in Episodically Consumed Foods in a Single-Replicate Study Design: EPIC Case Study.

In epidemiologic studies, measurement error in dietary variables often attenuates association between dietary intake and disease occurrence. To adjust for the attenuation caused by error in dietary intake, regression calibration is commonly used. To apply regression calibration, unbiased reference measurements are required. Short-term reference measurements for foods that are not consumed daily contain excess zeroes that pose challenges in the calibration model. We adapted two-part regression calibration model, initially developed for multiple replicates of reference measurements per individual to a single-replicate setting. We showed how to handle excess zero reference measurements by two-step modeling approach, how to explore heteroscedasticity in the consumed amount with variance-mean graph, how to explore nonlinearity with the generalized additive modeling (GAM) and the empirical logit approaches, and how to select covariates in the calibration model. The performance of two-part calibration model was compared with the one-part counterpart. We used vegetable intake and mortality data from European Prospective Investigation on Cancer and Nutrition (EPIC) study. In the EPIC, reference measurements were taken with 24-hour recalls. For each of the three vegetable subgroups assessed separately, correcting for error with an appropriately specified two-part calibration model resulted in about three fold increase in the strength of association with all-cause mortality, as measured by the log hazard ratio. Further found is that the standard way of including covariates in the calibration model can lead to over fitting the two-part calibration model. Moreover, the extent of adjusting for error is influenced by the number and forms of covariates in the calibration model. For episodically consumed foods, we advise researchers to pay special attention to response distribution, nonlinearity, and covariate inclusion in specifying the calibration model.

Vaccination of melanoma patients with Melan-A/Mart-1 peptide and Klebsiella outer membrane protein p40 as an adjuvant.

Toll-like receptor ligands are potentially useful adjuvants for the development of clinical T cell vaccination. Here we investigated the novel Toll-like receptor2 ligand P40, the outer membrane protein A derived from Klebsiella pneumoniae. Seventeen human leukocyte antigen-A*0201 positive stage III/IV melanoma patients were vaccinated with P40 and Melan-A/Mart-1 peptide subcutaneously in monthly intervals. Adverse reactions were mild-to-moderate. Fourteen patients received at least 8 vaccinations and were thus evaluable for clinical tumor and immune responses. Seven patients experienced progressive disease, whereas 2 patients had stable disease throughout the trial period, 1 of them with regression of multiple skin metastases. The remaining 5 patients had no measurable disease. Melan-A/Mart-1 specific CD8 T cells were analyzed ex vivo, with positive results in 6 of 14 evaluable patients. Increased percentages of T cells were found in three patients, memory/effector T cell differentiation in 4 patients, and a positive interferon-gamma Elispot assay in 1 patient. Antibody responses to P40 were observed in all patients. We conclude that vaccination with peptide and P40 was feasible and induced ex vivo detectable tumor antigen specific T cell responses in 6 of 14 patients with advanced melanoma.

Complete and long-lasting regression of disseminated multiple skin melanoma metastases under treatment with cyclooxygenase-2 inhibitor.

Experimental and clinical evidence indicates that non-steroidal anti-inflammatory drugs and cyclooxygenase-2 inhibitors may have anti-cancer activities. Here we report on a patient with a metastatic melanoma of the leg who experienced a complete and sustained regression of skin metastases upon continuous single treatment with the cyclooxygenase-2 inhibitor rofecoxib. Our observations indicate that the inhibition of cyclooxygenase-2 can lead to the regression of disseminated skin melanoma metastases, even after failure of chemotherapy.

Chlorophyll fluorescence parameters in populations of two legume trees: Stryphnodendron adstringens (Mart.) Coville (Mimosoideae) and Cassia ferruginea (Schrad.) Schrad. ex DC. (Caesalpinoideae)

The aim of this study was to investigate the photosynthetic performance in populations of two legume tree species, Stryphnodendron adstringens (Mimosoideae), typical from Cerrado, and Cassia ferruginea (Caesalpinoideae) from the Atlantic Rain Forest. The photosynthetic traits were assessed by measures of chlorophyll fluorescence in progenies of naturally pollinated plants from three populations of S. adstringens and a population of C. ferruginea. Plants of S. adstringens growing under similar conditions of C. ferruginea plants demanded higher light values for photosynthesis saturation, 600 µmol.m-2.s-1 and 350 µmol.m-2.s-1 respectively, and showed higher intrinsic photosynthetic efficiency of photosystem II, Fv/Fm of 0.814 versus 0.783 in C. ferruginea. The highest values of Fv/Fm observed in S. adstringens can explain the highest electron transport rates (ETR) obtained for this species. No significant differences were found among progenies from different C. ferruginea trees nor among populations of S. adstringens, and only in few cases, variation among progenies within populations were found for S. adstringens plants. The fact that fluorescence parameters distinguished species but not populations or most of progenies may be related to low intraspecific genetic variation of these chlorophyll fluorescence traits or due to lack of expression on genetic differences in plants under no stressful conditions.

Marolo (Annona crassiflora Mart.): a study of value chain and processing

This article aims to discuss the needs and problems of marolo value chain, as well as to evaluate the rehydration process of this fruit as a possibility of using it as a by-product during the interharvest growth periods. The study of the value chain included interviews with producers, handlers, and fruit and by-product sellers. In order to evaluate the rehydration process of this fruit, marolo was dehydrated using a conventional procedure and freeze-drying. The experiments were conducted in a completely randomized design and a triple factorial scheme (2 × 2 × 6). ANOVA was performed, followed by the Tukey's test (p < 0.05). Regression models were generated and adjusted for the time factor. The precariousness of the value chain of marolo was observed. The best procedure for marolo dehydration should be determined according to the intended use of the dehydrated product since the water-absorption capacity of the flour is higher and convective hot-air-drying is more effective in retaining soluble solids and reducing damage to the fruit. These results aim at contributing to the marolo value chain and to the preservation of native trees in the Brazilian savanna biome and can be used to analyze other underutilized crops.

Psychopathy and self-monitoring : additive and interactive effects on self-presentation tactics /

The purpose of this study was to determine the relative contributions of psychopathy and self-monitoring to the prediction of self-presentation tactics (behaviours that individuals use to manipulate their self-image). Psychopathy is composed of two main factors: Factor 1, which includes manipulativeness and shallow affect, and Factor 2, which includes irresponsibility and anti-social behaviours. Self-monitoring is a personality trait that distinguishes between those who adapt their behaviour to fit different social situations (high self-monitors) and those who behave as they feel regardless of social expectations (low selfmonitors). It was hypothesized that self-monitoring would moderate the relationship between psychopathy and self-presentation tactics. One hundred and forty-nine university students completed the Self-Monitoring Scale (Snyder, 1974), the Self-Report Psychopathy Scale - Version III (Paulhus et aI., in press), the Self-Presentation Tactics scale (Lee, S., et aI., 1999), the HEXACO-PI (a measure ofthe six major factors of personality; Lee, K., & Ashton, 2004), and six scenarios that were created as a supplementary measure of the selfpresentation tactics. Results of the hierarchical multiple regression analyses showed that self-monitoring did moderate the relationship between psychopathy and three of the selfpresentation tactics: apologies, disclaimers, and exemplification. Further, significant interactions were observed between Factor 1 and self-monitoring on apologies and the defensive tactics subscale, between Factor 2 and self-monitoring on self-handicapping, and between Factor 1 and Factor 2 on exemplification. Contrary to expectations, the main effect of self-monitoring was significant for the prediction of nine tactics, while psychopathy was significant for the prediction of seven tactics. This indicates that the role of these two personality traits in the explanation of self-presentation tactics tends to be additive in nature rather than interactive. In addition. Factor 2 alone did not account for a significant amount of variance in any of the tactics, while Factor 1 significantly predicted nine tactics. Results are discussed with regard to implications and possible directions for future research.

Aspects of spatial and habitat ecology of multiple Anopheles species (Diptera: Culicidae): malaria vectors in the highlands and foothills of Ecuador

The resurgence of malaria in highland regions of Africa, Oceania and recently in South America underlines the importance of the study of the ecology of highland mosquito vectors of malaria. Since the incidence of malaria is limited by the distribution of its vectors, the purpose of this PhD thesis was to examine aspects of the ecology of Anopheles mosquitoes in the Andes of Ecuador, South America. A historical literature and archival data review (Chapter 2) indicated that Anopheles pseudopunctipennis transmitted malaria in highland valleys of Ecuador prior to 1950, although it was eliminated through habitat removal and the use of chemical insecticides. Other anopheline species were previously limited to low-altitude regions, except in a few unconfirmed cases. A thorough larval collection effort (n=438 attempted collection sites) in all road-accessible parts of Ecuador except for the lowland Amazon basin was undertaken between 2008 - 2010 (Chapter 3). Larvae were identified morphologically and using molecular techniques (mitochondrial COl gene), and distribution maps indicated that all five species collected (Anopheles albimanus, An. pseudopunctipennis, Anopheles punctimacula, Anopheles oswaldoi s.l. and Anopheles eiseni) were more widespread throughout highland regions than previously recorded during the 1940s, with higher maximum altitudes for all except An. pseudopunctipennis (1541 m, 1930 m, 1906 m, 1233 m and 1873 m, respectively). During larval collections, to characterize species-specific larval habitat, a variety of abiotic and biotic habitat parameters were measured and compared between species-present and species-absent sites using chi-square tests and stepwise binary logistic regression analyses (Chapter 4). An. albimanus was significantly associated with permanent pools with sand substrates and An. pseudopunctipennis with gravel and boulder substrates. Both species were significantly associated with floating cyanobacterial mats and warmer temperatures, which may limit their presence in cooler highland regions. Anopheles punctimacula was collected more often than expected from algae-free, shaded pools with higher-than-average calculated dissolved oxygen. Anopheles oswaldoi s.l., the species occurring on the Amazonian side of the Andes, was associated with permanent, anthropogenic habitats such as roadside ditches and ponds. To address the hypothesis that human land use change is responsible for the emergence of multiple highland Anopheles species by creating larval habitat, common land uses in the western Andes were surveyed for standing water and potential larval habitat suitability (Chapter 5). Rivers and road edges provided large amounts of potentially suitable anopheline habitat in the western Andes, while cattle pasture also created potentially suitable habitat in irrigation canals and watering ponds. Other common land uses surveyed (banana farms, sugarcane plantations, mixed tree plantations, and empty lots) were usually established on steep slopes and had very little standing water present. Using distribution and larval habitat data, a GIS-based larval habitat distribution model for the common western species was constructed in ArcGIS v.l 0 (ESRI 2010) using derived data layers from field measurements and other sources (Chapter 6). The additive model predicted 76.4 - 97.9% of the field-observed collection localities of An. albimanus, An. pseudopunctipennis and An. punctimacula, although it could not accurately distinguish between species-absent and speciespresent sites due to its coarse scale. The model predicted distributional expansion and/or shift of one or more anopheline species into the following highland valleys with climate warming: Mira/Chota, Imbabura province, Tumbaco, Pichincha province, Pallatanga and Sibambe, Chimborazo province, and Yungilla, Azuay province. These valleys may serve as targeted sites of future monitoring to prevent highland epidemics of malaria. The human perceptions of malaria and mosquitoes in relation to land management practices were assessed through an interview-based survey (n=262) in both highlands and lowlands, of male and female land owners and managers of five property types (Chapter 7). Although respondents had a strong understanding of where the disease occurs in their own country and of the basic relationship among standing water, mosquitoes and malaria, about half of respondents in potential risk areas denied the current possibility of malaria infection on their own property. As well, about half of respondents with potential anopheline larval habitat did not report its presence, likely due to a highly specific definition of suitable mosquito habitat. Most respondents who are considered at risk of malaria currently use at least one type of mosquito bite prevention, most commonly bed nets. In conclusion, this interdisciplinary thesis examines the occurrence of Anopheles species in the lowland transition area and highlands in Ecuador, from a historic, geographic, ecological and sociological perspective.

Modelling and Analysis of Recurrent Event Data with Multiple Causes

This thesis Entitled “modelling and analysis of recurrent event data with multiple causes.Survival data is a term used for describing data that measures the time to occurrence of an event.In survival studies, the time to occurrence of an event is generally referred to as lifetime.Recurrent event data are commonly encountered in longitudinal studies when individuals are followed to observe the repeated occurrences of certain events. In many practical situations, individuals under study are exposed to the failure due to more than one causes and the eventual failure can be attributed to exactly one of these causes.The proposed model was useful in real life situations to study the effect of covariates on recurrences of certain events due to different causes.In Chapter 3, an additive hazards model for gap time distributions of recurrent event data with multiple causes was introduced. The parameter estimation and asymptotic properties were discussed .In Chapter 4, a shared frailty model for the analysis of bivariate competing risks data was presented and the estimation procedures for shared gamma frailty model, without covariates and with covariates, using EM algorithm were discussed. In Chapter 6, two nonparametric estimators for bivariate survivor function of paired recurrent event data were developed. The asymptotic properties of the estimators were studied. The proposed estimators were applied to a real life data set. Simulation studies were carried out to find the efficiency of the proposed estimators.

Sample size requirements for interval estimation of the strengthof association effect sizes in multiple regression analysis

Resumen tomado de la publicación

Random regressions models to describe the genetic variation of milk yield over multiple parities in Buffaloes

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Evidences of non-additive effects of multiple parasitoids on Diatraea saccharalis Fabr. (Lep., Crambidae) populations in sugarcane fields in Brazil

Biological control is a relatively benign method of pest control. However, considerable debate exists over whether multiple natural enemies often interact to produce additive or non-additive effects on their prey or host populations. Based on the large data set stored in the Sao Joao and Barra sugarcane mills (state of São Paulo, Brazil) regarding the programme of biological control of Diatraea saccharalis using the parasitoids Cotesia flavipes and tachinid flies, in the present study the author investigated whether the parasitoids released into sugarcane fields interfered significantly with the rate of parasitized D. saccharalis hosts. The author also observed whether there was an additive effect of releasing C. flavipes and tachinids on the rate of parasitized hosts, and looked for evidence of possible negative effects of the use of multiple parasitoid species in this biological control programme. Results showed that C. flavipes and the tachinids were concomitantly released in the Barra Mill, but not in the Sao Jao Mill. Furthermore, in the Barra Mill there was evidence that the parasitoids interacted because the percentage of parasitism did not increase after the release of either C. flavipes or tachinids. In the Sao Joao Mill, when both parasitoid species were released out of synchrony, both the percentage of parasitism by C. flavipes as well as that of the tachinids increased. When large numbers of tachinids were released in the Barra Mill, they caused a significant lower percentage of parasitism imposed by C. flavipes. The implications of the results as evidence of non-additive effects of C. flavipes plus tachinids on D. saccharalis populations are discussed.