The association between air pollution and blood pressure in traffic controllers in Santo Andre, Sao Paulo, Brazil

Background: Urban air pollutants are associated with cardiovascular events. Traffic controllers are at high risk for pollution exposure during outdoor work shifts. Objective: The purpose of this study was to evaluate the relationship between air pollution and systemic blood pressure in traffic controllers during their work shifts. Methods: This cross-sectional study enrolled 19 male traffic controllers from Santo Andre city (Sao Paulo, Brazil) who were 30-60 years old and exposed to ambient air during outdoor work shifts. Systolic and diastolic blood pressure readings were measured every 15 min by an Ambulatory Arterial Blood Pressure Monitoring device. Hourly measurements (lags of 0-5 h) and the moving averages (2-5 h) of particulate matter (PM(10)), ozone (O(3)) ambient concentrations and the acquired daily minimum temperature and humidity means from the Sao Paulo State Environmental Agency were correlated with both systolic and diastolic blood pressures. Statistical methods included descriptive analysis and linear mixed effect models adjusted for temperature, humidity, work periods and time of day. Results: Interquartile increases of PM(10) (33 mu g/m(3)) and O(3) (49 mu g/m(3)) levels were associated with increases in all arterial pressure parameters, ranging from 1.06 to 2.53 mmHg. PM(10) concentration was associated with early effects (lag 0), mainly on systolic blood pressure. However, O(3) was weakly associated most consistently with diastolic blood pressure and with late cumulative effects. Conclusions: Santo Andre traffic controllers presented higher blood pressure readings while working their outdoor shifts during periods of exposure to ambient pollutant fluctuations. However, PM(10) and O(3) induced cardiovascular effects demonstrated different time courses and end-point behaviors and probably acted through different mechanisms. (C) 2011 Elsevier Inc. All rights reserved.

ELECTROMYOGRAPHIC AMPLITUDE RATIO OF SERRATUS ANTERIOR AND UPPER TRAPEZIUS MUSCLES DURING MODIFIED PUSH-UPS AND BENCH PRESS EXERCISES

Imbalance and weakness of the serratus anterior and upper trapezius force couple have been described in patients with shoulder dysfunction. There is interest in identifying exercises that selectively activate these muscles and including it in rehabilitation protocols. This study aims to verify the UT/SA electromyographic (EMG) amplitude ratio, performed in different upper limb exercises and on two bases of support. Twelve healthy men were tested (average age = 22.8 +/- 3.1 years), and surface EMG was recorded from the upper trapezius and serratus anterior using single differential surface electrodes. Volunteers performed isometric contractions over a stable base of support and on a Swiss ball during the wall push-up (WP), bench press (BP), and push-up (PU) exercises. All SEMG data are reported as a percentage of root mean square or integral of linear envelope from the maximal value obtained in one of three maximal voluntary contractions for each muscle studied. A linear mixed-effect model was performed to compare UT/SA ratio values. The WP, BP, and PU exercises showed UT/SA ratio mean +/- SD values of 0.69 +/- 0.72, 0.14 +/- 0.12, and 0.39 +/- 0.37 for stable surfaces, respectively, whereas for unstable surfaces, the values were 0.73 +/- 0.67, 0.43 +/- 0.39, and 0.32 +/- 0.30. The results demonstrate that UT/SA ratio was influenced by the exercises and by the upper limb base of support. The practical application is to show that BP on a stable surface is the exercise preferred over WP and PU on either surfaces for serratus anterior muscle training in patients with imbalance between the UT/SA force couple or serratus anterior weakness.

Model specification tests in nonparametric stochastic regression models

In this paper, we consider testing for additivity in a class of nonparametric stochastic regression models. Two test statistics are constructed and their asymptotic distributions are established. We also conduct a small sample study for one of the test statistics through a simulated example. (C) 2002 Elsevier Science (USA).

Impact of academic exposure on health status of university students

OBJECTIVE: To assess the impact of academic life on health status of university students. METHODS: Longitudinal study including 154 undergraduate students from the Universidade de Aveiro, Portugal, with at least two years of follow-up observations. Sociodemographic and behavioral characteristics were collected using questionnaires. Students' weight, height, blood pressure, serum glucose, serum lipids and serum homocysteine levels were measured. Regression analysis was performed using linear mixed-effect models, allowing for random effects at the participant level. RESULTS: A higher rate of dyslipidemia (44.0% vs. 28.6%), overweight (16.3% vs. 12.5%) and smoking (19.3% vs. 0.0%) was found among students exposed to the academic life when compared to freshmen. Physical inactivity was about 80%. Total cholesterol, high density lipoprotein-cholesterol (HDL-C), triglycerides, systolic blood pressure, and physical activity levels were significantly associated with gender (p<0.001). Academic exposure was associated with increased low density lipoprotein-cholesterol (LDL-C) levels (about 1.12 times), and marginally with total cholesterol levels (p=0.041). CONCLUSIONS: High education level does not seem to have a protective effect favoring a healthier lifestyle and being enrolled in health-related areas does not seem either to positively affect students' behaviors. Increased risk factors for non-transmissible diseases in university students raise concerns about their well-being. These results should support the implementation of health promotion and prevention programs at universities.

Effect of the exposure to maternal smoking during pregnancy and childhood on the body mass index until adolescence

OBJECTIVE Investigate the effect of exposure to smoking during pregnancy and early childhood on changes in the body mass index (BMI) from birth to adolescence.METHODS A population-based cohort of children (0-5 years old) from Cuiabá, Midwest Brazil, was assessed in 1999-2000 (n = 2,405). Between 2009 and 2011, the cohort was re-evaluated. Information about birth weight was obtained from medical records, and exposure to smoking during pregnancy and childhood was assessed at the first interview. Linear mixed effects models were used to estimate the association between exposure to maternal smoking during pregnancy and preschool age, and the body mass index of children at birth, childhood and adolescence.RESULTS Only 11.3% of the mothers reported smoking during pregnancy, but most of them (78.2%) also smoked during early childhood. Among mothers who smoked only during pregnancy (n = 59), 97.7% had smoked only in the first trimester. The changes in body mass index at birth and in childhood were similar for children exposed and those not exposed to maternal smoking. However, from childhood to adolescence the rate of change in the body mass index was higher among those exposed only during pregnancy than among those who were not exposed.CONCLUSIONS Exposure to smoking only during pregnancy, especially in the first trimester, seems to affect changes in the body mass index until adolescence, supporting guidelines that recommend women of childbearing age to stop smoking.

Prognostic modelling of breast cancer patients: a benchmark of predictive models with external validation

Dissertação apresentada para obtenção do Grau de Doutor em Engenharia Electrotécnica e de Computadores – Sistemas Digitais e Percepcionais pela Universidade Nova de Lisboa, Faculdade de Ciências e Tecnologia

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.

Comparing the performance of geostatistical models with additional information from covariates for sewage plume characterization

In this work, kriging with covariates is used to model and map the spatial distribution of salinity measurements gathered by an autonomous underwater vehicle in a sea outfall monitoring campaign aiming to distinguish the effluent plume from the receiving waters and characterize its spatial variability in the vicinity of the discharge. Four different geostatistical linear models for salinity were assumed, where the distance to diffuser, the west-east positioning, and the south-north positioning were used as covariates. Sample variograms were fitted by the Mat`ern models using weighted least squares and maximum likelihood estimation methods as a way to detect eventual discrepancies. Typically, the maximum likelihood method estimated very low ranges which have limited the kriging process. So, at least for these data sets, weighted least squares showed to be the most appropriate estimation method for variogram fitting. The kriged maps show clearly the spatial variation of salinity, and it is possible to identify the effluent plume in the area studied. The results obtained show some guidelines for sewage monitoring if a geostatistical analysis of the data is in mind. It is important to treat properly the existence of anomalous values and to adopt a sampling strategy that includes transects parallel and perpendicular to the effluent dispersion.

Evolução, a curto prazo, do perfil do estado de nutrição no transplante hepático, em adultos : contributo para otimização da abordagem clínica e nutricional

RESUMO:Contexto: A avaliação do estado de nutrição do doente com indicação para transplante hepático (TH) deve ser abrangente, considerando o amplo espetro de situações clínicas e metabólicas. As alterações metabólicas relacionadas com a doença hepática podem limitar a aplicação de métodos de avaliação nutricional, subestimando a desnutrição. Após o TH, é expectável a reversão dos distúrbios metabólicos da doença hepática, pela melhoria da função do fígado. No entanto, algumas complicações metabólicas podem surgir após o TH, relacionadas com a má-nutrição, a desnervação hepática e o uso prolongado de imunossupressão, comprometendo os resultados clínicos a longo-prazo. A medição longitudinal e confiável do metabolismo energético e dos compartimentos corporais após o TH, avaliada em conjunto com fatores influentes no estado de nutrição, pode identificar precocemente situações de risco e otimizar e individualizar estratégias clínicas e nutricionais com vantagens no prognóstico. Objetivo: Avaliar longitudinalmente, a curto prazo, o estado de nutrição após o TH em doentes com insuficiência hepática por doença crónica e identificar os fatores, para além da cirurgia, que determinam diferentes evoluções do metabolismo energético e da composição corporal. Métodos: Foi estudada uma coorte de indivíduos com indicação para TH por doença hepática crónica, admitidos consecutivamente para TH ortotópico eletivo, durante 2 anos. Foram programados 3 momentos de avaliação: na última consulta pré-TH (T0), logo que adquirida autonomia respiratória e funcional após o TH (T1) e um mês após o TH (T2). Nesses momentos, foram medidos no mesmo dia: o suprimento nutricional por recordatório das últimas 24 horas, o estado de nutrição por Avaliação Subjetiva Global (ASG), o gasto energético em repouso (GER) por calorimetria indireta, a antropometria, a composição corporal por bioimpedância elétrica tetrapolar multifrequências e a força muscular por dinamometria de preensão palmar. O índice de massa magra (IMM) e a massa celular corporal (MCC) foram usados como indicadores do músculo esquelético e a percentagem de massa gorda (%MG) e o índice de massa gorda (IMG) como indicadores de adiposidade. O GER foi comparado com o estimado pelas fórmulas de Harris-Benedict para classificação do estado metabólico em:hipermetabolismo (GER medido >120% do GER estimado), normometabolismo (GER medido entre 80 e 120% do GER estimado) e hipometabolismo (GER medido <80% do GER estimado). Foi utilizada análise multivariável: por regressão logística, para identificar variáveis associadas à possibilidade (odds ratio – OR) de pertencer a cada grupo metabólico pré-TH; por regressão linear múltipla, para identificar variáveis associadas à variação dos compartimentos corporais no período pós-TH; e por modelos de efeitos mistos generalizados, para identificar variáveis associadas à evolução do GER e dos compartimentos corporais entre o período pré- e pós-TH. Resultados: Foram incluídos 56 indivíduos com idade, média (DP), 53,7 (8,5) anos, 87,5% do sexo masculino, 23,2% com doença hepática crónica de etiologia etanólica. Após o TH, em 60,7% indivíduos foi administrado regime imunossupressor baseado no tacrolimus. Os indivíduos foram avaliados [mediana (AIQ)] 90,5 (P25: 44,2; P75: 134,5) dias antes do TH (T0), 9,0 (P25: 7,0; P75: 12,0) dias após o TH (T1) e 36,0 (P25: 31,0; P75: 43,0) dias após o TH (T2). Após o TH houve melhoria significativa do estado de nutrição, com diminuição da prevalência de desnutrição classificada pela ASG (37,5% em T0, 16,1% em T2, p<0,001). Antes do TH, 41,1% dos indivíduos eram normometabólicos, 37,5% hipometabólicos e 21,4% hipermetabólicos. A possibilidade de pertencer a cada grupo metabólico pré-TH associou-se à: idade (OR=0,899, p=0,010) e desnutrição pela ASG (OR=5,038, p=0,015) para o grupo normometabólico; e índice de massa magra (IMM, OR=1,264, p=0,049) e etiologia viral da doença hepática (OR=8,297, p=0,019) para o grupo hipermetabólico. Não se obteve modelo múltiplo para o grupo de hipometabólico pré-TH, mas foram identificadas associações univariáveis com a história de toxicodependência (OR=0,282, p=0,047) e com a sarcopénia pré- TH (OR=8,000, p=0,040). Após o TH, houve normalização significativa e progressiva do estado metabólico, indicada pelo aumento da prevalência de normometabolismo (41,1% em T0, 57,1% em T2, p=0,040). Foram identificados diferentes perfis de evolução do GER após o TH, estratificado pelo estado metabólico pré-TH: no grupo hipometabólico pré-TH, o GER (Kcal) aumentou significativa e progressivamente (1030,6 em T0; 1436,1 em T1, p=0,001; 1659,2 em T2, p<0,001); no grupo hipermetabólico pré-TH o GER diminuiu significativa e progressivamente (2097,1 em T0; 1662,5 em T1, p=0,024; 1493,0 em T2, p<0.001); no grupo normometabólico não houve variações significativas. Os perfis de evolução do GER associaram-se com: peso corporal (β=9,6, p<0,001) e suprimento energético (β=13,6, p=0,005) na amostra total; com peso corporal (β=7,1, p=0,018) e contributo energético dos lípidos (β=18,9, p=0,003) no grupo hipometabólico pré-TH; e com peso corporal (β=14,1, p<0,001) e desnutrição pela ASG (β=-171,0, p=0,007) no grupo normometabólico pré-TH.Houve redução transitória dos compartimentos corporais entre T0 e T1, mas a maioria destes recuperou para valores semelhantes aos pré-TH. As exceções foram a água extracelular, que diminuiu entre T0 e T2 (média 18,2 L e 17,8 L, p=0,042), a massa gorda (média 25,1 Kg e 21,7 Kg, p<0,001) e o IMG (média 10,6 Kg.m-2 e 9,3 Kg.m-2, p<0,001) que diminuíram entre T1 e T2. Relativamente à evolução dos indicadores de músculo esquelético e adiposidade ao longo do estudo: a evolução do IMM associou-se com força de preensão palmar (β=0,06, p<0,001), creatininémia (β=2,28, p<0,001) e número total de fármacos administrados (β=-0,21, p<0,001); a evolução da MCC associou-se com força de preensão palmar (β=0,16, p<0,001), creatininémia (β=4,17, p=0,008) e número total de fármacos administrados (β=-0,46, p<0,001); a evolução da %MG associou-se com força de preensão palmar (β=-0,11, p=0,028), história de toxicodependência (β=-5,75, p=0,024), creatininémia (β=-5,91, p=0,004) e suprimento proteico (β=-0,06, p=0,001); a evolução do IMG associou-se com história de toxicodependência (β=- 2,64, p=0,019), creatininémia (β=-2,86, p<0,001) e suprimento proteico (β=-0,02, p<0,001). A variação relativa (%Δ) desses compartimentos corporais entre T1 e T2 indicou o impacto da terapêutica imunossupressora na composição corporal: o regime baseado na ciclosporina associou-se positivamente com a %Δ do IMM (β=23,76, p<0,001) e %Δ da MCC (β=26,58, p<0,001) e negativamente com a %Δ MG (β=-25,64, p<0,001) e %Δ do IMG (β=-25,62, p<0,001), relativamente ao regime baseado no tacrolimus. Os esteróides não influenciaram a evolução do GER nem com a dos compartimentos corporais. Conclusões: O estado de nutrição, avaliado por ASG, melhorou significativamente após o TH, traduzida pela diminuição da prevalência de desnutrição. O normometabolismo pré-TH foi prevalente e associou-se à menor idade e à desnutrição pré- TH. O hipometabolismo pré-TH associou-se à história de toxicodependência e à sarcopénia pré-TH. O hipermetabolismo pré-TH associou-se ao maior IMM e à etiologia viral da doença hepática. Após o TH, houve normalização progressiva do estado metabólico. Foram identificados três perfis de evolução do GER, associando-se com: peso corporal e suprimento energético na amostra total; peso corporal e contributo energético dos lípidos no grupo hipometabólico pré- TH; e peso corporal e desnutrição pela ASG no grupo normometabólico pré-TH. Foram identificados diferentes perfis de evolução da composição corporal após TH. A evolução do músculo esquelético associou-se positivamente com a força de preensão palmar e a creatininémia e negativamente com o número total de fármacos administrados. A evolução da adiposidade (%MG e IMG) associou-se inversamente com a história de toxicodependência, a creatininémia e o suprimento proteico; adicionalmente, a %MG associou-se inversamente com a força de preensão palmar. O regime baseado na ciclosporina associou-se independentemente com diminuição da adiposidade e aumento do músculo esquelético, comparativamente ao regime baseado no tacrolimus.---------------------------ABSTRACT:Background: The assessment of nutritional status in patients undergoing liver transplantation (LTx) should be comprehensive, accounting for the wide spectrum of the clinical and metabolic conditions. The metabolic disturbances related to liver disease may limit the precision and accuracy of traditional nutritional assessment methods underestimating the undernourishment. After LTx, it is expected that many metabolic derangements improve with the recovery of liver function. However, some metabolic complications arising after LTx, related to nutritional status, hepatic denervation, and prolonged immunosuppression, may compromise the longterm outcome. A reliable longitudinal assessment of both energy metabolism and body compartments after LTx, combined with assessments of other factors potentially affecting the nutritional status, may enable a better interpretation on the relationship between the metabolic and the nutritional status. These reliable assessments may precociously identify nutritional risk conditions and optimize and customize clinical and nutritional strategies improving the prognosis. Objective: To assess longitudinally the nutritional status shortly after orthotopic LTx in patients with chronic liver disease, and identify factors, beyond surgery, determining different energy metabolism and body composition profiles.Methods: A cohort of consecutive patients who underwent LTx due to chronic liver disease was studied within a period of two years. The assessments were performed in three occasions: at the last visit before LTx (T0), after surgery as soon as respiratory and functional autonomy was established (T1), and approximately one month after surgery (T2). On each occasion all assessments were performed on the same day, and included: the dietary assessment by 24- hour dietary recall, nutritional status by the Subjective Global Assessment (SGA), the resting energy expenditure (REE) by indirect calorimetry, anthropometry, body composition by multifrequency bioelectrical impedance analysis, and muscle strength by handgrip strength. Both the lean mass index (LMI) and body cell mass (BCM) were used as surrogates of skeletal muscle, and both the percentage of fat mass (%FM) and fat mass index (FMI) of adiposity. The REE was predicted according to the Harris and Benedict equation. Hypermetabolism was defined as a measured REE more than 120% of the predicted value; normometabolism as a measured REE within 80-120% of the predicted value; and hypometabolism as a measured REE less than 80% of the predicted value. Multiple regression analysis was used: by logistic regression to identify variables associated with odds of belong each pre-LTx metabolic groups; by linear multiple regression analysis to identify variables associated with body compartments relative variations (%Δ) in the post-LTx period; and by mixed effects models to identify variables associated with the REE and body compartments profiles pre- and post-LTx. Results: Fifty six patients with a mean (SD) of 53.7 (8.5) years of age were included, 87.5% were men and 23.2% with alcoholic liver disease. After LTx 60.7% individuals were assigned to tacrolimus-based immunosuppressive regimen. The patients were assessed at a median time (inter-quartil range) of 90.5 (P25 44.2; P75 134.5) days before LTx (T0), at a median time of 9.0 (P25 7.0; P75 12.0) (T1) and 36 (P25 31.0; P75 43.0) (T2) days after LTx. After LTx the nutritional status significantly improved: the SGA-undernourishment decreased from 37.5% (T0) to 16.1% (T2) (p<0.001). Before LTx, 41.1% patients were normometabolic, 37.5% hypometabolic, and 21.4% hypermetabolic. The predictors of each pre-LTx metabolic group were: age (OR=0.899, p=0.010) and SGA-undernourishment (OR=5.038, p=0.015) for the normometabolic group; and LMI (OR=1.264, p=0.049) and viral etiology of liver disease (OR=8.297, p=0.019) for the hypermetabolic group. No multiple model was found for the pre-LTx hypometabolic group, but univariate association was found with history of drug addiction (OR=0.282, p=0.047) and pre- LTx sarcopenia (OR=8.000, p=0.040). After LTx a significant normalization of the metabolic status occurred, indicated by the increase in the prevalence of normometabolic patients (from T0: 41.1% to T2: 57.1%, p=0.040). Different REE profiles were found with REE stratified by preoperative metabolic status: in the hypometabolic group a significant progressive increase in mean REE (Kcal) was observed (T0: 1030.6; T1: 1436.1, p=0.001; T2: 1659.2, p<0.001); in the hypermetabolic group, a significant progressive decrease in mean REE (Kcal) was observed (T0: 2097.1; T1: 1662.5, p=0.024; T2: 1493.0, p<0.001); and in the normometabolic group, no significant differences were found. The REE profiles were associated with: body weight (β- estimate=9.6, p<0.001) and energy intake (β-estimate=13.6, p=0.005) in the whole sample; with body weight (β-estimate=7.1, p=0.018) and %TEV from lipids (β-estimate=18.9, p=0.003) in the hypometabolic group; and with body weight (β-estimate=14.1, p<0.001), and SGAundernourishment (β-estimate=-171, p=0.007) in the normometabolic group. A transient decrease in most body compartments occurred from T0 to T1, with subsequent catch-up to similar preoperative values. Exceptions were the extracellular water, decreasing from T0 to T2 (mean 18.2 L to 17.8 L, p=0.042), the fat mass (mean 25.1 Kg to 21.7 Kg, p<0.001) and FMI (mean 10.6 Kg.m-2 to 9.3 Kg.m-2, p<0.001), decreasing from T1 to T2. Significant predictors of skeletal muscle and adiposity profiles were found: LMI evolution was associated with handgrip strength (β-estimate=0.06, p<0.001), serum creatinine (β- estimate=2.28, p<0.001) and number of medications (β-estimate=-0.21, p<0.001); BCM evolution was associated with handgrip strength (β-estimate=0.16, p<0.001), serum creatinine (β-estimate=4.17, p<0.001) and number of medications (β-estimate=-0.46, p<0.001); the %FM evolution was associated with handgrip strength (β-estimate=-0.11, p=0.028), history of drug addiction (β-estimate=-5.75, p=0.024), serum creatinine (β-estimate=-5.91, p=0.004) and protein intake (β-estimate=-0.06, p=0.001); and FMI evolution was associated with history of drug addiction (β-estimate=-2.64, p=0.019), serum creatinine (β-estimate=-2.86, p<0.001) and protein intake (β-estimate=-0.02, p<0.001). The %Δ of the aforementioned body compartments from T1 to T2 indicated the influence of immunosuppressive agents on body composition: the cyclosporine-based regimen, compared with tacrolimus-based regimen, was positively associated with %Δ LMI (β-estimate=23.76, p<0.001) and %Δ BCM (β- estimate=26.58, p<0.001), and inversely associated with %Δ FM (β-estimate=-25.64, p<0.001) and %Δ FMI (β-estimate=-25.62, p<0.001). No significant changes in REE or body composition were observed associated with dose or duration of steroid therapy. Conclusions: The SGA-assessed nutritional status improved shortly after LTx, with significant decrease in prevalence undernourished individuals. XXI Preoperative normometabolism was prevalent and was associated with younger age and SGAundernourishment before LTx. Preoperative hypometabolism was associated with history of drug addiction and pre-LTx sarcopenia. Preoperative hypermetabolism was associated with higher LMI and viral etiology of liver disease. A significant normalization of the metabolic status was observed after LTx. The REE profiles were positively predicted by body weight and energy intake in the whole sample, by body weight and percentage of energy intake from lipids in the preoperative hypometabolic patients, and by body weight and SGA–undernourishment in the preoperative normometabolic patients. Different body composition profiles were found after LTx. Skeletal muscle profile was positively associated with handgrip strength and serum creatinine, and inversely with the number of medications. The adiposity profile was inversely associated with history of drug addiction, serum creatinine and protein intake. Additionally, the %FM evolution was inversely associated with handgrip strength. The cyclosporine-based regimen, compared with tacrolimus-based regimen, was independently associated with skeletal muscle increase and adiposity decrease.

Blood Pressure Variation Throughout Pregnancy According to Early Gestational BMI: A Brazilian Cohort

Background: The maternal cardiovascular system undergoes progressive adaptations throughout pregnancy, causing blood pressure fluctuations. However, no consensus has been established on its normal variation in uncomplicated pregnancies. Objective: To describe the variation in systolic blood pressure (SBP) and diastolic blood pressure (DBP) levels during pregnancy according to early pregnancy body mass index (BMI). Methods: SBP and DBP were measured during the first, second and third trimesters and at 30-45 days postpartum in a prospective cohort of 189 women aged 20-40 years. BMI (kg/m2) was measured up to the 13th gestational week and classified as normal-weight (<25.0) or excessive weight (≥25.0). Longitudinal linear mixed-effects models were used for statistical analysis. Results: A decrease in SBP and DBP was observed from the first to the second trimester (βSBP=-0.394; 95%CI: -0.600- -0.188 and βDBP=-0.617; 95%CI: -0.780- -0.454), as was an increase in SBP and DBP up to 30-45 postpartum days (βSBP=0.010; 95%CI: 0.006-0.014 and βDBP=0.015; 95%CI: 0.012-0.018). Women with excessive weight at early pregnancy showed higher mean SBP in all gestational trimesters, and higher mean DBP in the first and third trimesters. Excessive early pregnancy BMI was positively associated with prospective changes in SBP (βSBP=7.055; 95%CI: 4.499-9.610) and in DBP (βDBP=3.201; 95%CI: 1.136-5.266). Conclusion: SBP and DBP decreased from the first to the second trimester and then increased up to the postpartum period. Women with excessive early pregnancy BMI had higher SBP and DBP than their normal-weight counterparts throughout pregnancy, but not in the postpartum period.

A priori ratemaking using bivariate poisson regression models

In automobile insurance, it is useful to achieve a priori ratemaking by resorting to gene- ralized linear models, and here the Poisson regression model constitutes the most widely accepted basis. However, insurance companies distinguish between claims with or without bodily injuries, or claims with full or partial liability of the insured driver. This paper exa- mines an a priori ratemaking procedure when including two di®erent types of claim. When assuming independence between claim types, the premium can be obtained by summing the premiums for each type of guarantee and is dependent on the rating factors chosen. If the independence assumption is relaxed, then it is unclear as to how the tari® system might be a®ected. In order to answer this question, bivariate Poisson regression models, suitable for paired count data exhibiting correlation, are introduced. It is shown that the usual independence assumption is unrealistic here. These models are applied to an automobile insurance claims database containing 80,994 contracts belonging to a Spanish insurance company. Finally, the consequences for pure and loaded premiums when the independence assumption is relaxed by using a bivariate Poisson regression model are analysed.

The ghost of past species occurrence: improving species distribution models for presence-only data

1. Model-based approaches have been used increasingly in conservation biology over recent years. Species presence data used for predictive species distribution modelling are abundant in natural history collections, whereas reliable absence data are sparse, most notably for vagrant species such as butterflies and snakes. As predictive methods such as generalized linear models (GLM) require absence data, various strategies have been proposed to select pseudo-absence data. However, only a few studies exist that compare different approaches to generating these pseudo-absence data. 2. Natural history collection data are usually available for long periods of time (decades or even centuries), thus allowing historical considerations. However, this historical dimension has rarely been assessed in studies of species distribution, although there is great potential for understanding current patterns, i.e. the past is the key to the present. 3. We used GLM to model the distributions of three 'target' butterfly species, Melitaea didyma, Coenonympha tullia and Maculinea teleius, in Switzerland. We developed and compared four strategies for defining pools of pseudo-absence data and applied them to natural history collection data from the last 10, 30 and 100 years. Pools included: (i) sites without target species records; (ii) sites where butterfly species other than the target species were present; (iii) sites without butterfly species but with habitat characteristics similar to those required by the target species; and (iv) a combination of the second and third strategies. Models were evaluated and compared by the total deviance explained, the maximized Kappa and the area under the curve (AUC). 4. Among the four strategies, model performance was best for strategy 3. Contrary to expectations, strategy 2 resulted in even lower model performance compared with models with pseudo-absence data simulated totally at random (strategy 1). 5. Independent of the strategy model, performance was enhanced when sites with historical species presence data were not considered as pseudo-absence data. Therefore, the combination of strategy 3 with species records from the last 100 years achieved the highest model performance. 6. Synthesis and applications. The protection of suitable habitat for species survival or reintroduction in rapidly changing landscapes is a high priority among conservationists. Model-based approaches offer planning authorities the possibility of delimiting priority areas for species detection or habitat protection. The performance of these models can be enhanced by fitting them with pseudo-absence data relying on large archives of natural history collection species presence data rather than using randomly sampled pseudo-absence data.

Evaluating the ability of habitat suitability models to predict species presences.

Models predicting species spatial distribution are increasingly applied to wildlife management issues, emphasising the need for reliable methods to evaluate the accuracy of their predictions. As many available datasets (e.g. museums, herbariums, atlas) do not provide reliable information about species absences, several presence-only based analyses have been developed. However, methods to evaluate the accuracy of their predictions are few and have never been validated. The aim of this paper is to compare existing and new presenceonly evaluators to usual presence/absence measures. We use a reliable, diverse, presence/absence dataset of 114 plant species to test how common presence/absence indices (Kappa, MaxKappa, AUC, adjusted D-2) compare to presenceonly measures (AVI, CVI, Boyce index) for evaluating generalised linear models (GLM). Moreover we propose a new, threshold-independent evaluator, which we call "continuous Boyce index". All indices were implemented in the B10MAPPER software. We show that the presence-only evaluators are fairly correlated (p > 0.7) to the presence/absence ones. The Boyce indices are closer to AUC than to MaxKappa and are fairly insensitive to species prevalence. In addition, the Boyce indices provide predicted-toexpected ratio curves that offer further insights into the model quality: robustness, habitat suitability resolution and deviation from randomness. This information helps reclassifying predicted maps into meaningful habitat suitability classes. The continuous Boyce index is thus both a complement to usual evaluation of presence/absence models and a reliable measure of presence-only based predictions.

Time-series regression models to study the short-term effects of environmental factors on health

Time series regression models are especially suitable in epidemiology for evaluating short-term effects of time-varying exposures on health. The problem is that potential for confounding in time series regression is very high. Thus, it is important that trend and seasonality are properly accounted for. Our paper reviews the statistical models commonly used in time-series regression methods, specially allowing for serial correlation, make them potentially useful for selected epidemiological purposes. In particular, we discuss the use of time-series regression for counts using a wide range Generalised Linear Models as well as Generalised Additive Models. In addition, recently critical points in using statistical software for GAM were stressed, and reanalyses of time series data on air pollution and health were performed in order to update already published. Applications are offered through an example on the relationship between asthma emergency admissions and photochemical air pollutants

Correcting Measurement Error Bias in Interaction Models with Small Samples

Several methods have been suggested to estimate non-linear models with interaction terms in the presence of measurement error. Structural equation models eliminate measurement error bias, but require large samples. Ordinary least squares regression on summated scales, regression on factor scores and partial least squares are appropriate for small samples but do not correct measurement error bias. Two stage least squares regression does correct measurement error bias but the results strongly depend on the instrumental variable choice. This article discusses the old disattenuated regression method as an alternative for correcting measurement error in small samples. The method is extended to the case of interaction terms and is illustrated on a model that examines the interaction effect of innovation and style of use of budgets on business performance. Alternative reliability estimates that can be used to disattenuate the estimates are discussed. A comparison is made with the alternative methods. Methods that do not correct for measurement error bias perform very similarly and considerably worse than disattenuated regression