923 resultados para Quantile regressions
Resumo:
A finalidade deste estudo foi descrever a sobrecarga e o desconforto emocional dos cuidadores de idosos. Estudo epidemiológico e transversal conduzido em 2009 com 124 cuidadores residentes na comunidade de Ribeirão Preto-SP, por meio dos instrumentos: Escala de Sobrecarga de Zarit e Self-Reporting Questionaire (SRQ-20) para o cuidador. A análise dos dados foi realizada no aplicativo SPSS, 15.0, de forma descritiva, univariada (tabelas de frequência) e bivariada (tabelas de contingência para variáveis qualitativas). Os cuidadores, 85,6% do sexo feminino, média de 56,5 anos, utilizaram, em média, 12,4 horas diárias para o cuidar e 57,6% dos cuidadores apresentaram de leve a moderada sobrecarga. Dependência funcional do idoso, sexo do cuidador e tempo em horas para o cuidado, foram preditores da sobrecarga (p<0,05). Encontrou-se, também, que a sobrecarga é fator de risco para desconforto emocional (p<0,05). Cabe aos enfermeiros utilizarem protocolos de avaliação, com base nos fatores de risco, para prevenir a sobrecarga.
Resumo:
OBJETIVO: Identificar e quantificar a influência dos fatores socioeconômicos sobre os padrões alimentares. MÉTODOS: Estudo transversal de base populacional com amostra de 1.136 crianças e adolescentes de 7 a 14 anos de idade, de ambos os sexos, matriculados na rede pública de Salvador (BA), Brasil. O consumo alimentar foi medido por meio do questionário qualitativo de frequência alimentar. Os padrões de consumo foram identificados por meio de análise de componentes principais. Para o estudo da influência dos indicadores socioeconômicos na conformação dos padrões alimentares, foram utilizados modelos de regressão quantílica. RESULTADOS: Os padrões alimentares extraídos foram classificados em padrão obesogênico e padrão tradicional. Nos modelos de regressão quantílica, ajustados por faixa etária e por sexo, o menor grau de instrução materna esteve associado negativamente, em níveis significantes, na maioria dos percentis, ao consumo de alimentos que integram o padrão obesogênico. A baixa renda associou-se negativamente aos maiores percentis (p>95). Os dados indicam não haver influência dos indicadores socioeconômicos sobre o consumo de alimentos que integram o padrão tradicional. CONCLUSÃO: Conclui-se que há influência dos fatores socioeconômicos na adesão ao padrão obesogênico de consumo. Esse conjunto de resultados requer a atenção dos gestores públicos para a identificação de um padrão de consumo ocidental, visualizado amplamente nos estudos em que se avaliam padrões de consumo adotados na atualidade pela população brasileira - sobretudo por crianças e adolescentes -, caracterizados por englobar componentes alimentares de risco para as doenças crônicas não transmissíveis.
Resumo:
Insulin-like growth factor type 1 (IGF1) is a mediator of growth hormone (GH) action, and therefore, IGF1 is a candidate gene for recombinant human GH (rhGH) pharmacogenetics. Lower serum IGF1 levels were found in adults homozygous for 19 cytosine-adenosine (CA) repeats in the IGF1 promoter. The aim of this study was to evaluate the influence of (CA)n IGF1 polymorphism, alone or in combination with GH receptor (GHR)-exon 3 and -202 A/C insulin-like growth factor binding protein-3 (IGFBP3) polymorphisms, on the growth response to rhGH therapy in GH-deficient (GHD) patients. Eighty-four severe GHD patients were genotyped for (CA) n IGF1, -202 A/C IGFBP3 and GHR-exon 3 polymorphisms. Multiple linear regressions were performed to estimate the effect of each genotype, after adjustment for other influential factors. We assessed the influence of genotypes on the first year growth velocity (1st y GV) (n = 84) and adult height standard deviation score (SDS) adjusted for target-height SDS (AH-TH SDS) after rhGH therapy (n = 37). Homozygosity for the IGF1 19CA repeat allele was negatively correlated with 1st y GV (P = 0.03) and AH-TH SDS (P = 0.002) in multiple linear regression analysis. In conjunction with clinical factors, IGF1 and IGFBP3 genotypes explain 29% of the 1st y GV variability, whereas IGF1 and GHR polymorphisms explain 59% of final height-target-height SDS variability. We conclude that homozygosity for IGF1 (CA) 19 allele is associated with less favorable short-and long-term growth outcomes after rhGH treatment in patients with severe GHD. Furthermore, this polymorphism exhibits a non-additive interaction with -202 A/C IGFBP3 genotype on the 1st y GV and with GHR-exon 3 genotype on adult height. The Pharmacogenomics Journal (2012) 12, 439-445; doi:10.1038/tpj.2011.13; published online 5 April 2011
Resumo:
Abstract Background Lower respiratory tract infection (LRTI) is a major cause of pediatric morbidity and mortality, especially among non-affluent communities. In this study we determine the impact of respiratory viruses and how viral co-detections/infections can affect clinical LRTI severity in children in a hospital setting. Methods Patients younger than 3 years of age admitted to a tertiary hospital in Brazil during the months of high prevalence of respiratory viruses had samples collected from nasopharyngeal aspiration. These samples were tested for 13 different respiratory viruses through real-time PCR (rt-PCR). Patients were followed during hospitalization, and clinical data and population characteristics were collected during that period and at discharge to evaluate severity markers, especially length of hospital stay and oxygen use. Univariate regression analyses identified potential risk factors and multivariate logistic regressions were used to determine the impact of specific viral detections as well as viral co-detections in relation to clinical outcomes. Results We analyzed 260 episodes of LRTI with a viral detection rate of 85% (n = 222). Co-detection was observed in 65% of all virus-positive episodes. The most prevalent virus was Respiratory Syncytial Virus (RSV) (54%), followed by Human Metapneumovirus (hMPV) (32%) and Human Rhinovirus (HRV) (21%). In the multivariate models, infants with co-detection of HRV + RSV stayed 4.5 extra days (p = 0.004), when compared to infants without the co-detection. The same trends were observed for the outcome of days of supplemental oxygen use. Conclusions Although RSV remains as the main cause of LRTI in infants our study indicates an increase in the length of hospital stay and oxygen use in infants with HRV detected by RT-PCR compared to those without HRV. Moreover, one can speculate that when HRV is detected simultaneously with RSV there is an additive effect that may be reflected in more severe clinical outcome. Also, our study identified a significant number of children infected by recently identified viruses, such as hMPV and Human Bocavirus (HBov), and this is a novel finding for poor communities from developing countries.
HPV clearance in postpartum period of HIV-positive and negative women: a prospective follow-up study
Resumo:
Abstract Background HPV persistence is a key determinant of cervical carcinogenesis. The influence of postpartum on HPV clearance has been debated. This study aimed to assess HPV clearance in later pregnancy and postpartum among HIV-positive and negative women. Methods We conducted a follow-up study with 151 HPV-positive women coinfected with HIV, in 2007–2010. After baseline assessment, all women were retested for HPV infection using PCR in later pregnancy and after delivery. Multivariable logistic regressions assessed the putative association of covariates with HPV status in between each one of the successive visits. Results Seventy-one women (47%) have eliminated HPV between the baseline visit and their second or third visits. HIV-positive women took a significantly longer time (7.0 ± 3.8 months) to clear HPV, compared to those not infected by HIV (5.9 ± 3.0 months). HPV clearance was significantly more likely to take place after delivery than during pregnancy (84.5% x 15.5%). Conclusions Both HIV-positive and negative women presented a significant reduction in HPV infection during the postpartum period. HIV-positive status was found to be associated with a longer period of time to clear HPV infection in pregnant women.
Resumo:
Abstract Background Regardless the regulatory function of microRNAs (miRNA), their differential expression pattern has been used to define miRNA signatures and to disclose disease biomarkers. To address the question of whether patients presenting the different types of diabetes mellitus could be distinguished on the basis of their miRNA and mRNA expression profiling, we obtained peripheral blood mononuclear cell (PBMC) RNAs from 7 type 1 (T1D), 7 type 2 (T2D), and 6 gestational diabetes (GDM) patients, which were hybridized to Agilent miRNA and mRNA microarrays. Data quantification and quality control were obtained using the Feature Extraction software, and data distribution was normalized using quantile function implemented in the Aroma light package. Differentially expressed miRNAs/mRNAs were identified using Rank products, comparing T1DxGDM, T2DxGDM and T1DxT2D. Hierarchical clustering was performed using the average linkage criterion with Pearson uncentered distance as metrics. Results The use of the same microarrays platform permitted the identification of sets of shared or specific miRNAs/mRNA interaction for each type of diabetes. Nine miRNAs (hsa-miR-126, hsa-miR-1307, hsa-miR-142-3p, hsa-miR-142-5p, hsa-miR-144, hsa-miR-199a-5p, hsa-miR-27a, hsa-miR-29b, and hsa-miR-342-3p) were shared among T1D, T2D and GDM, and additional specific miRNAs were identified for T1D (20 miRNAs), T2D (14) and GDM (19) patients. ROC curves allowed the identification of specific and relevant (greater AUC values) miRNAs for each type of diabetes, including: i) hsa-miR-1274a, hsa-miR-1274b and hsa-let-7f for T1D; ii) hsa-miR-222, hsa-miR-30e and hsa-miR-140-3p for T2D, and iii) hsa-miR-181a and hsa-miR-1268 for GDM. Many of these miRNAs targeted mRNAs associated with diabetes pathogenesis. Conclusions These results indicate that PBMC can be used as reporter cells to characterize the miRNA expression profiling disclosed by the different diabetes mellitus manifestations. Shared miRNAs may characterize diabetes as a metabolic and inflammatory disorder, whereas specific miRNAs may represent biological markers for each type of diabetes, deserving further attention.
Resumo:
Measurements on the growth process and placental development of the embryo and fetuses of Cavia porcellus were carried out using ultrasonography. Embryo, fetus, and placenta were monitored from Day 15 after mating day to the end of gestation. Based on linear and quadratic regressions, the following morphometric analysis showed a good indicator of the gestational age: placental diameter, biparietal diameter, renal length, and crown rump. The embryonic cardiac beat was first detected at an average of 22.5 days. The placental diameter showed constant increase from beginning of gestation then remained to term and presented a quadratic correlation with gestational age (r2 = 0.89). Mean placental diameter at the end of pregnancy was 3.5 ± 0.23 cm. By Day 30, it was possible to measure biparietal diameter, which followed a linear pattern of increase up to the end of gestation (r2 = 0.95). Mean biparietal diameter in the end of pregnancy was 1.94 ± 0.03 cm. Kidneys were firstly observed on Day 35 as hyperechoic structures without the distinction of medullar and cortical layers, thus the regression model equation between kidney length and gestational age presents a quadratic relationship (r2 = 0.7). The crown rump presented a simple linear growth, starting from 15 days of gestation, displaying a high correlation with the gestational age (r2 = 0.9). The offspring were born after an average gestation of 61.3 days. In this study, we conclude that biparietal diameter, placental diameter, and crown rump are adequate predictive parameters of gestational age in guinea pigs because they present high correlation index.
Resumo:
Linear dispersal systems, such as coastal habitats, are well suited for phylogeographic studies because of their low spatial complexity compared to three dimensional habitats. Widely distributed coastal plant species additionally show azonal and often essentially continuous distributions. These properties, firstly, make it easier to reconstruct historical distributions of coastal plants and, secondly, make it more likely that present distributions contain both Quaternary refugia and recently colonized areas. Taken together this makes it easier to formulate phylogeographic hypotheses. This work investigated the phylogeography of Cakile maritima and Eryngium maritimum, two species growing in sandy habitats along the north Atlantic Ocean and the Mediterranean Sea coasts on two different spatial scales using AFLP data. The genetic structure of these species was investigated by sampling single individuals along most of their distributions from Turkey to south Sweden. On a regional scale the population genetic structure of both species was also studied in detail in the Bosporus and Dardanelles straits, the Strait of Gibraltar and along a continuous stretch of dunes in western France. Additionally, populations of C. maritima were investigated in the Baltic Sea/Kattegat/North Sea area. Over the complete sampling range the species show both differences and similarities in their genetic structure. In the Mediterranean Sea, both species contain Aegean Sea/Black Sea and west Mediterranean clusters. Cakile maritima additionally shows a clustering of Ionian Sea/Adriatic Sea collections. Further, both species show a subdivision of Atlantic Ocean/North Sea/Baltic Sea material from Mediterranean. Within the Atlantic Ocean group, C. maritima from the Baltic Sea and the most northern Atlantic localities form an additional cluster while no such substructure was found in E. maritimum. In all three instances where population genetic investigations of both species were performed in the same area, the results showed almost complete congruency of spatial genetic patterns. In the Aegean/Black Sea/Marmara region a subdivision of populations into a Black Sea, a Sea of Marmara and an Aegean Sea group is shared by both species. In addition the Sea of Marmara populations are more close to the Aegean Sea populations than they are to the Black Sea populations in both cases. Populations from the Atlantic side of the Strait of Gibraltar are differentiated from those on the Mediterranean side in both species, a pattern that confirms the results of the wide scale study. Along the dunes of West France no clear genetic structure could be detected in any of the species. Additionally, the results from the Baltic Sea/North Sea populations of C. maritima did not reveal any geographical genetic pattern. It is postulated that the many congruencies between the species are mainly due to a predominantly sea water mediated seed dispersal in both species and their shared sandy habitat. The results are compared to hypothetical distributions for the last glacial maximum based on species specific temperature requirements. It is argued that in both species the geographical borders of the clusters in the Mediterranean area were not affected by quaternary temperature changes and that the Aegean/Black Sea/Marmara cluster, and possibly the Ionian Sea/Adriatic Sea cluster in C. maritima, is the result of sea currents that isolate these basins from the rest of the sampled areas. The genetic gap in the Strait of Gibraltar between Atlantic Ocean and Mediterranean Sea populations in both species is also explained in terms of sea currents. The existence of three subgroups corresponding to the Aegean Sea, Black Sea and Sea of Marmara basins is suggested to have arisen due to geographical isolation during periods of global sea regressions in the glacials. The population genetic evidence was inconclusive regarding the Baltic Sea cluster of C. Maritima from the wide scale study. The results of this study are very similar to those of an investigation of three other coastal plant species over a similar range. This suggests that the phylo-geographic patterns of widespread coastal plants may be more predictable than those of other terrestrial plants.
Resumo:
Die Untersuchungen umfassen die Periode 1981 – 2000 und basieren hauptsächlich auf Daten des Deutschen Wetterdienstes (DWD). Relativwerte der Globalstrahlung beziehen sich auf die Rayleigh-Atmosphäre. Das Regressionsmodell nach Angström ermöglicht die Erweiterung des Meßnetzes. In linearer und nichtlinearer Regression und Korrelation ist die Globalstrahlung entweder abhängige (Sonnenscheindauer, Bewölkung) oder unabhängige Variable (Lufttemperatur, Bodentemperatur). Ihre Intensität in Abhängigkeit von Großwetterlagen, Großwettertypen und Luftmassen wird diskutiert. Diesbezüglich werden mit der Linearen Diskriminanzanalyse ähnliche Großwetterlagen und Stationen in signifikant unterschiedenen Gruppen zusammengefaßt, getrennt nach Sommer- und Winterhalbjahr. Abhängig von der Zeit betrachtet, enthalten Globalstrahlung, direkte und diffuse Sonnenstrahlung, Lufttemperatur, Bewölkung und Niederschlag signifikante zyklische Variationen, die gegebenenfalls klimatologisch relevant sind. Weiteren Aufschluß ergeben deshalb die Zeitreihenanalysen. Autokorrelation-Spektralanalysen (ASA) der genannten Variablen werden in integrierten Spektren dargestellt. Hinweise auf die zeitliche Konstanz signifikanter Varianzmaxima enthalten die Spektren der dynamischen (gleitenden) ASA.
Resumo:
Eine Voraussetzung für die Entwicklung neuer immunmodulatorischer Therapieverfahren ist die Kenntnis immunogener Tumorantigene, die von tumorreaktiven T-Zellen erkannt werden. In der vorliegenden Arbeit wurden tumorreaktive CD8+ zytotoxische T-Lymphozyten (CTL, cytotoxic T-lymphocytes) aus dem Blut eines HLA (human leukocyte antigen)-kompatiblen Fremdspenders generiert. Methodisch wurden hierzu CD8-selektionierte periphere Blutlymphozyten repetitiv mit der klarzelligen Nierenzellkarzinomlinie MZ1851-RCC (RCC, renal cell carcinoma) in einer allogenen gemischten Lymphozyten-Tumorzell Kultur (MLTC, mixed lymphocyte tumor cell culture) stimuliert. Aus den Responderlymphozyten wurden mit Hilfe des Grenzverdünnungsverfahrens klonale zytotoxische T-Zellen generiert und expandiert. Die CTL-Klone wurden anschließend phänotypisch mittels Durchflußzytometrie sowie funktionell mittels HLA-Antikörper-Blockadeexperimenten und Kreuzreaktivitätstests detailliert charakterisiert. Dabei konnte gezeigt werden, daß aus dem Blut eines allogenen gesunden Spenders CD8+ T-Zellen isoliert werden können, welche Reaktivität gegen Nierenzellkarzinome (NZK) aufweisen und über verschiedene HLA-Klasse-I-Allele restringiert sind. Die von den einzelnen CTL-Klonen erkannten Zielstrukturen zeigten entweder ubiquitäre (z.B. HLA-Cw*0704-reaktiver CTL-Klon E77) oder eine tumorspezifische (z.B. HLA-B*0702-restringierter CTL-Klon A4) Gewebeexpression. Zur Identifizierung der natürlich prozessierten Peptidliganden wurden die HLA-B/C-Allele unter Verwendung des monoklonalen Antikörpers B123.2 aus einem zuvor hergestellten Detergenslysat der Nierenzellkarzinomlinie MZ1851-RCC immunchromatographisch aufgereinigt. Aus den so isolierten HLA-Peptid-Komplexen wurden die tumorassoziierten Peptidliganden nach Säureeluation und Filtration abgespalten und über eine „reverse phase“-HPLC (high performance liquid chromatography) fraktioniert. Die Überprüfung der einzelnen HPLC-Fraktionen auf Bioaktivität erfolgte mit den korrespondierenden CTL-Klonen in 51Cr-Zytotoxizitätstests. Dabei wurde eine HPLC-Fraktion identifiziert, die die lytische Funktion des HLA-B*0702-restringierten CTL-Klons A4 auslösen konnte. Die bioaktive HPLC-Fraktion wurde dazu durch eine zweite (second dimension) Kapillar-Flüssigkeitschromatographie (Cap-LC, capillar liquid chromatography) in Subfraktionen geringerer Komplexität aufgetrennt und die darin enthaltenen Peptidepitope durch das MALDI-TOF/TOF (matrix assisted laser desorption/ionization- time of flight/time of flight)-Analyseverfahren sequenziert. Innerhalb dieser HPLC-Fraktion wurden eine Vielzahl von HLA-B/C-assoziierten Peptidliganden erfolgreich sequenziert, was die Effektivität dieser Verfahrenstechnik zur Identifizierung natürlich prozessierter HLA-Klasse-I-bindender Peptide unter Beweis stellt. Leider war es mit dieser Methode bisher nicht möglich, das von CTL-Klon A4 detektierte Peptidepitop zu sequenzieren. Dies liegt möglicherweise in der unzureichenden Konzentration des Peptidepitops in der bioaktiven HPLC-Fraktion begründet. In Folgearbeiten soll nun mit erhöhter Probenmenge beziehungsweise verbesserter Analytik der erneute Versuch unternommen werden, das Zielantigen des CTL-Klons A4 zu identifizieren. Die Kenntnis von Antigenen, die tumorspezifisch exprimiert und von CD8+ CTL aus gesunden Spendern erkannt werden, eröffnet neue therapeutische Möglichkeiten, das spezifische Immunsystem des Stammzellspenders nach allogener Blutstammzelltransplantation gezielt zur Steigerung von Tumorabstoßungsreaktionen (z.B. durch Vakzinierung oder adoptivem T-Zelltransfer) zu nutzen.
Resumo:
This doctoral thesis aims at contributing to the literature on transition economies focusing on the Russian Federations and in particular on regional income convergence and fertility patterns. The first two chapter deal with the issue of income convergence across regions. Chapter 1 provides an historical-institutional analysis of the period between the late years of the Soviet Union and the last decade of economic growth and a presentation of the sample with a description of gross regional product composition, agrarian or industrial vocation, labor. Chapter 2 contributes to the literature on exploratory spatial data analysis with a application to a panel of 77 regions in the period 1994-2008. It provides an analysis of spatial patterns and it extends the theoretical framework of growth regressions controlling for spatial correlation and heterogeneity. Chapter 3 analyses the national demographic patterns since 1960 and provides a review of the policies on maternity leave and family benefits. Data sources are the Statistical Yearbooks of USSR, the Statistical Yearbooks of the Russian Soviet Federative Socialist Republic and the Demographic Yearbooks of Russia. Chapter 4 analyses the demographic patterns in light of the theoretical framework of the Becker model, the Second Demographic Transition and an economic-crisis argument. With national data from 1960, the theoretically issue of the pro or countercyclical relation between income and fertility is graphically analyzed and discussed, together with female employment and education. With regional data after 1994 different panel data models are tested. Individual level data from the Russian Longitudinal Monitoring Survey are employed using the logit model. Chapter 5 employs data from the Generations and Gender Survey by UNECE to focus on postponement and second births intentions. Postponement is studied through cohort analysis of mean maternal age at first birth, while the methodology used for second birth intentions is the ordered logit model.
Resumo:
„Risikomaße in der Finanzmathematik“ Der Value-at -Risk (VaR) ist ein Risikomaß, dessen Verwendung von der Bankenaufsicht gefordert wird. Der Vorteil des VaR liegt – als Quantil der Ertrags- oder Verlustverteilung - vor allem in seiner einfachen Interpretierbarkeit. Nachteilig ist, dass der linke Rand der Wahrscheinlichkeitsverteilung nicht beachtet wird. Darüber hinaus ist die Berechnung des VaR schwierig, da Quantile nicht additiv sind. Der größte Nachteil des VaR ist in der fehlenden Subadditivität zu sehen. Deswegen werden Alternativen wie Expected Shortfall untersucht. In dieser Arbeit werden zunächst finanzielle Risikomaße eingeführt und einige ihre grundlegenden Eigenschaften festgehalten. Wir beschäftigen uns mit verschiedenen parametrischen und nichtparametrischen Methoden zur Ermittlung des VaR, unter anderen mit ihren Vorteilen und Nachteilen. Des Weiteren beschäftigen wir uns mit parametrischen und nichtparametrischen Schätzern vom VaR in diskreter Zeit. Wir stellen Portfoliooptimierungsprobleme im Black Scholes Modell mit beschränktem VaR und mit beschränkter Varianz vor. Der Vorteil des erstens Ansatzes gegenüber dem zweiten wird hier erläutert. Wir lösen Nutzenoptimierungsprobleme in Bezug auf das Endvermögen mit beschränktem VaR und mit beschränkter Varianz. VaR sagt nichts über den darüber hinausgehenden Verlust aus, während dieser von Expected Shortfall berücksichtigt wird. Deswegen verwenden wir hier den Expected Shortfall anstelle des von Emmer, Korn und Klüppelberg (2001) betrachteten Risikomaßes VaR für die Optimierung des Portfolios im Black Scholes Modell.
Resumo:
Form und Gestalt kraniofazialer Strukturen sind primär beeinflusst durch die inhärente Integration der unterschiedlichsten Funktionssysteme und externer selektiver Einflüsse. Die Variabilität der Schädel-Morphe ist ein Indikator für solche Einflussfaktoren und damit ein idealer Gegenstand für vergleichende Analysen morphogenetischer Formbildung. Zur Ermittlung morphologisch-adaptiver Trends und Muster wurden sowohl Hypothesen zur morphologischen Differenziertheit als auch zu Korrelationen zwischen modularen Schädelkompartimenten (fazial, neurokranial, basikranial) untersucht. Zusätzlich wurden aus Schichtröntgenaufnahmen (CT) virtuelle Modelle rekonstruiert, welche die Interpretation der statistischen Befunde unterstützen sollten. Zur Berechnung der Gestaltunterschiede wurden mittels eines mechanischen Gelenkarm-Messgerätes (MicroScribe-G2) max. 85 ektokraniale Messpunkte (Landmarks) bzw. dreidimensionale Koordinaten an ca. 520 Schädeln von fünf rezenten Gattungen der Überfamilie Hominoidea (Hylobates, Pongo, Gorilla, Pan und Homo) akquiriert. Aus dem Datensatz wurden geometrische Störfaktoren (Größe, Translation, Rotation) mathematisch eliminiert und die verbleibenden Residuale bzw. ‚Gestalt-Variablen‘ diversen multivariat-statistischen Verfahren unterzogen (Faktoren, Cluster-, Regressions- und Korrelationsanalysen sowie statistische Tests). Die angewandten Methoden erhalten die geometrische Information der Untersuchungsobjekte über alle Analyseschritte hinweg und werden unter der Bezeichnung „Geometric Morphometrics (GMM)“ als aktueller Ansatz der Morphometrie zusammengefasst. Für die unterschiedlichen Fragestellungen wurden spezifische Datensätze generiert. Es konnten diverse morphologische Trends und adaptive Muster mit Hilfe der Synthese statistischer Methoden und computer-basierter Rekonstruktionen aus den generierten Datensätzen ermittelt werden. Außerdem war es möglich, präzise zu rekonstruieren, welche kranialen Strukturen innerhalb der Stichprobe miteinander wechselwirken, einzigartige Variabilitäten repräsentieren oder eher homogen gestaltet sind. Die vorliegenden Befunde lassen erkennen, dass Fazial- und Neurokranium am stärksten miteinander korrelieren, während das Basikranium geringe Abhängigkeiten in Bezug auf Gesichts- oder Hirnschädelveränderungen zeigte. Das Basikranium erweist sich zudem bei den nicht-menschlichen Hominoidea und über alle Analysen hinweg als konservative und evolutiv-persistente Struktur mit dem geringsten Veränderungs-Potential. Juvenile Individuen zeigen eine hohe Affinität zueinander und zu Formen mit einem kleinem Gesichts- und großem Hirnschädel. Während das Kranium des rezenten Menschen primär von Enkephalisation und fazialer Retraktion (Orthognathisierung) dominiert ist und somit eine einzigartige Gestalt aufweist, zeigt sich der Kauapparat als maßgeblich formbildendes Kompartiment bei den nicht-menschlichen Formen. Die Verbindung von GMM mit den interaktiven Möglichkeiten computergenerierter Modelle erwies sich als valides Werkzeug zur Erfassung der aufgeworfenen Fragestellungen. Die Interpretation der Befunde ist durch massive Interkorrelationen der untersuchten Strukturen und der statistisch-mathematischen Prozeduren als hoch komplex zu kennzeichnen. Die Studie präsentiert einen innovativen Ansatz der modernen Morphometrie, welcher für zukünftige Untersuchungen im Bereich der kraniofazialen Gestaltanalyse ausgebaut werden könnte. Dabei verspricht die Verknüpfung mit ‚klassischen’ und modernen Zugängen (z. B. Molekularbiologie) gesteigerte Erkenntnismöglichkeiten für künftige morphometrische Fragestellungen.
Resumo:
The aim of this thesis is to apply multilevel regression model in context of household surveys. Hierarchical structure in this type of data is characterized by many small groups. In last years comparative and multilevel analysis in the field of perceived health have grown in size. The purpose of this thesis is to develop a multilevel analysis with three level of hierarchy for Physical Component Summary outcome to: evaluate magnitude of within and between variance at each level (individual, household and municipality); explore which covariates affect on perceived physical health at each level; compare model-based and design-based approach in order to establish informativeness of sampling design; estimate a quantile regression for hierarchical data. The target population are the Italian residents aged 18 years and older. Our study shows a high degree of homogeneity within level 1 units belonging from the same group, with an intraclass correlation of 27% in a level-2 null model. Almost all variance is explained by level 1 covariates. In fact, in our model the explanatory variables having more impact on the outcome are disability, unable to work, age and chronic diseases (18 pathologies). An additional analysis are performed by using novel procedure of analysis :"Linear Quantile Mixed Model", named "Multilevel Linear Quantile Regression", estimate. This give us the possibility to describe more generally the conditional distribution of the response through the estimation of its quantiles, while accounting for the dependence among the observations. This has represented a great advantage of our models with respect to classic multilevel regression. The median regression with random effects reveals to be more efficient than the mean regression in representation of the outcome central tendency. A more detailed analysis of the conditional distribution of the response on other quantiles highlighted a differential effect of some covariate along the distribution.
Resumo:
During my Doctoral study I researched about the remote detection of canopy N concentration in forest stands, its potentials and problems, under many overlapping perspectives. The study consisted of three parts. In S. Rossore 2000 dataset analysis, I tested regressions between N concentration and NIR reflectances derived from different sources (field samples, airborne and satellite sensors). The analysis was further expanded using a larger dataset acquired in year 2009 as part of a new campaign funded by the ESA. In both cases, a good correlation was observed between Landsat NIR, using both TM (2009) and ETM+ (2000) imagery, and N concentration measured by a CHN elemental analyzer. Concerning airborne sensors I did not obtain the same good results, mainly because of the large FOV of the two instruments, and to the anisotropy of vegetation reflectance. We also tested the relation between ground based ASD measures and nitrogen concentration, obtaining really good results. Thus, I decided to expand my study to the regional level, focusing only on field and satellite measures. I analyzed a large dataset for the whole of Catalonia, Spain; MODIS imagery was used, in consideration of its spectral characteristics and despite its rather poor spatial resolution. Also in this case a regression between nitrogen concentration and reflectances was found, but not so good as in previous experiences. Moreover, vegetation type was found to play an important role in the observed relationship. We concluded that MODIS is not the most suitable satellite sensor in realities like Italy and Catalonia, which present a patchy and inhomogeneous vegetation cover; so it could be utilized for the parameterization of eco-physiological and biogeochemical models, but not for really local nitrogen estimate. Thus multispectral sensors similar to Landsat Thematic Mapper, with better spatial resolution, could be the most appropriate sensors to estimate N concentration.