An administrative claims model for profiling hospital 30-day mortality rates for pneumonia patients

Background

Outcome measures for patients hospitalized with pneumonia may complement process measures in characterizing quality of care. We sought to develop and validate a hierarchical regression model using Medicare claims data that produces hospital-level, risk-standardized 30-day mortality rates useful for public reporting for patients hospitalized with pneumonia.

Methodology/Principal Findings

Retrospective study of fee-for-service Medicare beneficiaries age 66 years and older with a principal discharge diagnosis of pneumonia. Candidate risk-adjustment variables included patient demographics, administrative diagnosis codes from the index hospitalization, and all inpatient and outpatient encounters from the year before admission. The model derivation cohort included 224,608 pneumonia cases admitted to 4,664 hospitals in 2000, and validation cohorts included cases from each of years 1998–2003. We compared model-derived state-level standardized mortality estimates with medical record-derived state-level standardized mortality estimates using data from the Medicare National Pneumonia Project on 50,858 patients hospitalized from 1998–2001. The final model included 31 variables and had an area under the Receiver Operating Characteristic curve of 0.72. In each administrative claims validation cohort, model fit was similar to the derivation cohort. The distribution of standardized mortality rates among hospitals ranged from 13.0% to 23.7%, with 25^th, 50^th, and 75^th percentiles of 16.5%, 17.4%, and 18.3%, respectively. Comparing model-derived risk-standardized state mortality rates with medical record-derived estimates, the correlation coefficient was 0.86 (Standard Error = 0.032).

Conclusions/Significance

An administrative claims-based model for profiling hospitals for pneumonia mortality performs consistently over several years and produces hospital estimates close to those using a medical record model.     

Breeding value accuracy estimates for growth traits using random regression and multi-trait models in Nelore cattle

We quantified the potential increase in accuracy of expected breeding value for weights of Nelore cattle, from birth to mature age, using multi-trait and random regression models on Legendre polynomials and B-spline functions. A total of 87,712 weight records from 8144 females were used, recorded every three months from birth to mature age from the Nelore Brazil Program. For random regression analyses, all female weight records from birth to eight years of age (data set I) were considered. From this general data set, a subset was created (data set II), which included only nine weight records: at birth, weaning, 365 and 550 days of age, and 2, 3, 4, 5, and 6 years of age. Data set II was analyzed using random regression and multi-trait models. The model of analysis included the contemporary group as fixed effects and age of dam as a linear and quadratic covariable. In the random regression analyses, average growth trends were modeled using a cubic regression on orthogonal polynomials of age. Residual variances were modeled by a step function with five classes. Legendre polynomials of fourth and sixth order were utilized to model the direct genetic and animal permanent environmental effects, respectively, while third-order Legendre polynomials were considered for maternal genetic and maternal permanent environmental effects. Quadratic polynomials were applied to model all random effects in random regression models on B-spline functions. Direct genetic and animal permanent environmental effects were modeled using three segments or five coefficients, and genetic maternal and maternal permanent environmental effects were modeled with one segment or three coefficients in the random regression models on B-spline functions. For both data sets (I and II), animals ranked differently according to expected breeding value obtained by random regression or multi-trait models. With random regression models, the highest gains in accuracy were obtained at ages with a low number of weight records. The results indicate that random regression models provide more accurate expected breeding values than the traditionally finite multi-trait models. Thus, higher genetic responses are expected for beef cattle growth traits by replacing a multi-trait model with random regression models for genetic evaluation. B-spline functions could be applied as an alternative to Legendre polynomials to model covariance functions for weights from birth to mature age.     

Modeling of crossbred cattle growth, comparison between cubic and piecewise random regression models

Two analyses, cubic and piecewise random regression, were conducted to model growth of crossbred cattle from birth to about two years of age, investigating the ability of a piecewise procedure to fit growth traits without the complications of the cubic model. During a four-year period (1994-1997) of the Australian "Southern Crossbreeding Project", mature Hereford cows (N = 581) were mated to 97 sires of Angus, Belgian Blue, Hereford, Jersey, Limousin, South Devon, and Wagyu breeds, resulting in 1141 steers and heifers born over four years. Data included 13 (for steers) and eight (for heifers) live body weight measurements, made approximately every 50 days from birth until slaughter. The mixed model included fixed effects of sex, sire breed, age (linear, quadratic and cubic), and their interactions between sex and sire breed with age. Random effects were sire, dam, management (birth location, year, post-weaning groups), and permanent environmental effects and for each of these when possible, their interactions with linear, quadratic and cubic growth. In both models, body weights of all breeds increased over pre-weaning period, held fairly steady (slightly flattening) over the dry season then increased again towards the end of the feedlot period. The number of estimated parameters for the cubic model was 22 while for the piecewise model it was 32. It was concluded that the piecewise model was very similar to the cubic model in the fit to the data; with the piecewise model being marginally better. The piecewise model seems to fit the data better at the end of the growth period.     

Using Structured Additive Regression Models to Estimate Risk Factors of Malaria: Analysis of 2010 Malawi Malaria Indicator Survey Data

Background

After years of implementing Roll Back Malaria (RBM) interventions, the changing landscape of malaria in terms of risk factors and spatial pattern has not been fully investigated. This paper uses the 2010 malaria indicator survey data to investigate if known malaria risk factors remain relevant after many years of interventions.

Methods

We adopted a structured additive logistic regression model that allowed for spatial correlation, to more realistically estimate malaria risk factors. Our model included child and household level covariates, as well as climatic and environmental factors. Continuous variables were modelled by assuming second order random walk priors, while spatial correlation was specified as a Markov random field prior, with fixed effects assigned diffuse priors. Inference was fully Bayesian resulting in an under five malaria risk map for Malawi.

Results

Malaria risk increased with increasing age of the child. With respect to socio-economic factors, the greater the household wealth, the lower the malaria prevalence. A general decline in malaria risk was observed as altitude increased. Minimum temperatures and average total rainfall in the three months preceding the survey did not show a strong association with disease risk.

Conclusions

The structured additive regression model offered a flexible extension to standard regression models by enabling simultaneous modelling of possible nonlinear effects of continuous covariates, spatial correlation and heterogeneity, while estimating usual fixed effects of categorical and continuous observed variables. Our results confirmed that malaria epidemiology is a complex interaction of biotic and abiotic factors, both at the individual, household and community level and that risk factors are still relevant many years after extensive implementation of RBM activities.     

Quality of life and related factors among HIV-positive spouses from serodiscordant couples under antiretroviral therapy in Henan Province, China

Objective

To describe the quality of life and related factors in HIV-positive spouses undergoing ART from discordant couples.

Methods

A cross-sectional study was conducted among 1,009 HIV-positive spouses from serodiscordant couples in Zhumadian, Henan Province, between October 1, 2008 and March 31, 2009. HIV-positive spouses were interviewed by local health professionals. Quality of life was evaluated by WHOQOL (Chinese Version). A multiple linear regression model was used to analyze the related factors.

Results

The majority of subjects were female (56.39%), had received a high school education (44%), were of Han ethnicity (98.41%), and were farmers (90.09%); the median time period of receiving ART was 3.92 years. The physical, psychological, social, and environmental QOL scores of the subjects were 12.91±1.95, 12.35±1.80, 13.96±2.43, and 12.45±1.91 respectively. The multiple linear regression model identified the physical domain related factors to be CD4 count, educational level, and occupation; psychological domain related factors include age, educational level, and reported STD symptom; social domain related factors included education level; and environmental domain related factors included education level, reported STD symptoms, and occupation.

Conclusion

Being younger, a farmer, having a lower level of education, a reported STD symptom, or lower CD4 count, could decrease one''s quality of life, suggesting that the use of blanket ART programs alone may not necessarily improve quality of life. Subjects received lower scores in the psychological domain, suggesting that psychological intervention may also need to be strengthened.     

Climate factors associated with historic spruce beetle (Coleoptera: Curculionidae) outbreaks in Utah and Colorado

This study investigated relationships between climate and historic spruce beetle, Dendroctonus rufipennis Kirby (Coleoptera: Curculionidae), outbreaks in northern and southeastern Utah and western Colorado between 1905 and 1996. A chronology of outbreak years was constructed from historic records, research papers, newspapers, and other sources of information. Historic climate data for the region included annual and mean monthly temperature and precipitation, in addition to Palmer drought severity index (PDSI) values estimated from tree rings. Classification and regression tree analysis (CART) was used to identify those climate factors most important for predicting historic spruce beetle outbreaks. The factors identified by the best CART model included mean December temperature, mean September temperature 1 yr before outbreak years, the mean estimated PDSI value of the 5-yr period before outbreak years, and mean October precipitation. The resulting model correctly classified nonoutbreak and outbreak years 67 and 70% of the time, respectively.     

Persistence of accuracy of genome-wide breeding values over generations when including a polygenic effect

Background

When estimating marker effects in genomic selection, estimates of marker effects may simply act as a proxy for pedigree, i.e. their effect may partially be attributed to their association with superior parents and not be linked to any causative QTL. Hence, these markers mainly explain polygenic effects rather than QTL effects. However, if a polygenic effect is included in a Bayesian model, it is expected that the estimated effect of these markers will be more persistent over generations without having to re-estimate the marker effects every generation and will result in increased accuracy and reduced bias.

Methods

Genomic selection using the Bayesian method, ''BayesB'' was evaluated for different marker densities when a polygenic effect is included (GWpEBV) and not included (GWEBV) in the model. Linkage disequilibrium and a mutation drift balance were obtained by simulating a population with a Ne of 100 over 1,000 generations.

Results

Accuracy of selection was slightly higher for the model including a polygenic effect than for the model not including a polygenic effect whatever the marker density. The accuracy decreased in later generations, and this reduction was stronger for lower marker densities. However, no significant difference in accuracy was observed between the two models. The linear regression of TBV on GWEBV and GWpEBV was used as a measure of bias. The regression coefficient was more stable over generations when a polygenic effect was included in the model, and was always between 0.98 and 1.00 for the highest marker density. The regression coefficient decreased more quickly with decreasing marker density.

Conclusions

Including a polygenic effect had no impact on the selection accuracy, but showed reduced bias, which is especially important when estimates of genome-wide markers are used to estimate breeding values over more than one generation.     

Temporal water stress effects on nodulation,nitrogen accumulation and growth of soybean

The objective of this study, carried out over 2 years, was to evaluate the effect of soil properties on the response of maize (Zea mays L.) to zinc applications and relate these properties to soil test Zn for predicting the Zn status of soils considering the effect of environmental conditions. The relative yield, expressed as an index of crop response, was related through multiple regression to CEC (or clay), electrical conductivity (or exchangeable Na), and bulk density consistently throughout the two year period that included one relatively wet, cool and cloudy growing season when variations in relative yield were explained also by 0.5M NaHCO₃-extractable-P and organic C. A procedure is presented to establish limits for the soil propeties and soil-test-extractable-Zn and to meaningfully combine them into a model to predict soil Zn status. A model that combined soil test Zn, texture and electrical conductivity was satisfactory for the purpose of prediction and for adoption for soil testing on a routine basis. The suggested approach may be suitable for designing models with soil properties associated with crop responses to micronutrients in other situations. Deceased 22 September 1988     

Functional alterations in cardiac muscle after medium- or long-term simulated weightlessness and related cellular mechanisms

A serial work on effects of simulated weightlessness on function and structure of cardiac muscle was started in our laboratory several years ago. In our previous papers, a long-term tail-suspension rat model modified by us, and its validity as a simulation of microgravity effects on the cardiovascular system have been reported. In the present paper, we will focus primarily on the nature, time course, and mechanism of functional alterations in rat cardiac muscle during both 4 wk of tail-suspension by our technique and 2 wk of recovery. Besides, some findings concerning changes after 13-wk tail-suspension and the regression during recovery for 3 wk are also included.     

Rheumatoid arthritis patients are not at increased risk for 30-day cardiovascular events,infections, or mortality after total joint arthroplasty

Introduction

Serious infection, cardiovascular disease, and mortality are increased in rheumatoid arthritis (RA). Whether RA affects the risk for these complications after total joint arthroplasty (TJA) is unknown, we hypothesize that it does. We compared the occurrence of 30-day postoperative complications and mortality in a large cohort of RA and osteoarthritis (OA) patients undergoing hip or knee TJA.

Methods

Analyses included 7-year data from the Veterans Affairs Surgical Quality Improvement Program. The 30-day complications were compared by diagnosis by using logistic regression, and long-term mortality was examined by using Cox proportional hazards regression. All analyses were adjusted for age, sex, and clustering by surgical site. Additional covariates included sociodemographics, comorbidities, health behaviors, and operative risk factors.

Results

The 34,524 patients (839 RA, 33,685 OA) underwent knee (65.9%) or hip TJA. Patients were 95.7% men with a mean (SD) age of 64.4 (10.7) years and had 3,764 deaths over a mean follow-up of 3.7 (2.3) years. Compared with OA patients, those with RA were significantly more likely to require a return to the operating room (odds ratio (OR), 1.45 (95% CI, 1.08 to 1.94), but had similar rates of 30-day postoperative infection, OR 1.02 (0.72 to 1.47), cardiovascular events, OR 0.69 (0.37 to 1.28), and mortality, OR 0.94 (0.38 to 2.33). RA was associated with a significantly higher long-term mortality; hazard ratio (HR), 1.22 (1.00 to 1.49).

Conclusion

In this study of US veterans, RA patients were not at an increased risk for short-term mortality or other major complications after TJA, although they returned to the operating room more often and had increased long-term mortality.     

Predicting Unprecedented Dengue Outbreak Using Imported Cases and Climatic Factors in Guangzhou, 2014

Introduction

Dengue is endemic in more than 100 countries, mainly in tropical and subtropical regions, and the incidence has increased 30-fold in the past 50 years. The situation of dengue in China has become more and more severe, with an unprecedented dengue outbreak hitting south China in 2014. Building a dengue early warning system is therefore urgent and necessary for timely and effective response.

Methodology and Principal Findings

In the study we developed a time series Poisson multivariate regression model using imported dengue cases, local minimum temperature and accumulative precipitation to predict the dengue occurrence in four districts of Guangzhou, China. The time series data were decomposed into seasonal, trend and remainder components using a seasonal-trend decomposition procedure based on loess (STL). The time lag of climatic factors included in the model was chosen based on Spearman correlation analysis. Autocorrelation, seasonality and long-term trend were controlled in the model. A best model was selected and validated using Generalized Cross Validation (GCV) score and residual test. The data from March 2006 to December 2012 were used to develop the model while the data from January 2013 to September 2014 were employed to validate the model. Time series Poisson model showed that imported cases in the previous month, minimum temperature in the previous month and accumulative precipitation with three month lags could project the dengue outbreaks occurred in 2013 and 2014 after controlling the autocorrelation, seasonality and long-term trend.

Conclusions

Together with the sole transmission vector Aedes albopictus, imported cases, monthly minimum temperature and monthly accumulative precipitation may be used to develop a low-cost effective early warning system.     

Fusing Data Mining,Machine Learning and Traditional Statistics to Detect Biomarkers Associated with Depression

Background

Atheoretical large-scale data mining techniques using machine learning algorithms have promise in the analysis of large epidemiological datasets. This study illustrates the use of a hybrid methodology for variable selection that took account of missing data and complex survey design to identify key biomarkers associated with depression from a large epidemiological study.

Methods

The study used a three-step methodology amalgamating multiple imputation, a machine learning boosted regression algorithm and logistic regression, to identify key biomarkers associated with depression in the National Health and Nutrition Examination Study (2009–2010). Depression was measured using the Patient Health Questionnaire-9 and 67 biomarkers were analysed. Covariates in this study included gender, age, race, smoking, food security, Poverty Income Ratio, Body Mass Index, physical activity, alcohol use, medical conditions and medications. The final imputed weighted multiple logistic regression model included possible confounders and moderators.

Results

After the creation of 20 imputation data sets from multiple chained regression sequences, machine learning boosted regression initially identified 21 biomarkers associated with depression. Using traditional logistic regression methods, including controlling for possible confounders and moderators, a final set of three biomarkers were selected. The final three biomarkers from the novel hybrid variable selection methodology were red cell distribution width (OR 1.15; 95% CI 1.01, 1.30), serum glucose (OR 1.01; 95% CI 1.00, 1.01) and total bilirubin (OR 0.12; 95% CI 0.05, 0.28). Significant interactions were found between total bilirubin with Mexican American/Hispanic group (p = 0.016), and current smokers (p<0.001).

Conclusion

The systematic use of a hybrid methodology for variable selection, fusing data mining techniques using a machine learning algorithm with traditional statistical modelling, accounted for missing data and complex survey sampling methodology and was demonstrated to be a useful tool for detecting three biomarkers associated with depression for future hypothesis generation: red cell distribution width, serum glucose and total bilirubin.     

Hip Fracture in the Elderly: A Re-Analysis of the EPIDOS Study with Causal Bayesian Networks

Objectives

Hip fractures commonly result in permanent disability, institutionalization or death in elderly. Existing hip-fracture predicting tools are underused in clinical practice, partly due to their lack of intuitive interpretation. By use of a graphical layer, Bayesian network models could increase the attractiveness of fracture prediction tools. Our aim was to study the potential contribution of a causal Bayesian network in this clinical setting. A logistic regression was performed as a standard control approach to check the robustness of the causal Bayesian network approach.

Setting

EPIDOS is a multicenter study, conducted in an ambulatory care setting in five French cities between 1992 and 1996 and updated in 2010. The study included 7598 women aged 75 years or older, in which fractures were assessed quarterly during 4 years. A causal Bayesian network and a logistic regression were performed on EPIDOS data to describe major variables involved in hip fractures occurrences.

Results

Both models had similar association estimations and predictive performances. They detected gait speed and mineral bone density as variables the most involved in the fracture process. The causal Bayesian network showed that gait speed and bone mineral density were directly connected to fracture and seem to mediate the influence of all the other variables included in our model. The logistic regression approach detected multiple interactions involving psychotropic drug use, age and bone mineral density.

Conclusion

Both approaches retrieved similar variables as predictors of hip fractures. However, Bayesian network highlighted the whole web of relation between the variables involved in the analysis, suggesting a possible mechanism leading to hip fracture. According to the latter results, intervention focusing concomitantly on gait speed and bone mineral density may be necessary for an optimal prevention of hip fracture occurrence in elderly people.     

Relationship of Leptin Concentration to Gender,Menopause, Age,Diabetes, and Fat Mass in African Americans

This investigation was designed to determine the relationship of leptin concentration to gender, sex hormones, menopause, age, diabetes, and fat mass in African Americans. Participants included 101 African Americans, 38 men (mean age, 34. 2 ± 7. 4 years), 29 age-matched premenopausal women (mean age, 32. 6 ± 3. 7 years), and 36 postmenopausal women (mean age, 57. 8 ± 5. 9 years). The women were not taking exogenous sex hormones, and 12 subjects were diabetic. Percent body fat was calculated with the Siri formula, fat mass (FM) was calculated as weight x percent body fat, and Fat-free mass (FFM) was calculated as weight minus FM. Fasting plasma was assayed for leptin, estradiol, free testosterone, glucose, and insulin concentrations. The nondiabetics had an oral glucose tolerance test (OGTT). The diabetics compared with the non-diabetics had a higher central fat index (^P=0. 04) but otherwise were similar to nondiabetics in all parameters measured. Body mass index, percent body fat, and FM were greater in women than men (p<0. 001). Leptin concentrations in men, premenopausal, and postmenopausal women were: 7. 51 ± 8. 5, 33. 9 ± 17. 3, 31. 4 ± 22. 3 ng/mL. Leptin/FM x 100 in the three groups were: 28. 9 ± 16. 1, 98. 65 ± 44. 9, 77. 1 ± 44. 5 ng/mL/kg. The gender difference in leptin concentration and leptin/FM was significant (p<0. 001), but the difference between premenopausal and postmenopausal women was not. In each group, weight, percent body fat, and FM were highly correlated with leptin concentration. Multiple regression analyses with leptin concentration as the dependent variable and age, diabetic status, percent body fat, weight, FM, FFM, estradiol, and free testosterone concentrations as independent variables demonstrated that the determinants of leptin concentration in men was weight only (R=0. 83,p<0. 001), in premenopausal women it was FM only (R=0. 57,P<0. 001), and in postmenopausal women it was weight only (R=0. 67, p<0. 001). With diabetics excluded, the multiple regression analysis was repeated with fasting insulin concentration and the area under the insulin curve during the OGTT included as independent variables. The results for this multiple regression analyses were the same as the first. Therefore, leptin concentration in African Americans is determined by gender and fat mass. Menopause, age, and diabetes do not affect leptin concentration.     

Progeroid syndrome with facial teleangiectatic erythema, posterior subcapsular cataracts, calcification of basal ganglia and atrium septum defect type 2

In this report we present the long-term follow-up findings in a young female born to consanguineous parents with the unique association of (1) a progeroid syndrome, (2) facial dysmorphism with relative microcephaly, triangular face, retrognathism and skin erythema, (3) bilateral posterior cataracts, (4) basal ganglia calcifications and (5) atrium septum defect type 2. Intelligence is borderline. Clinical evolution after normal puberty was positive with regression of the facial erythematous changes. Over the years differential diagnosis included progeria, hypohidrotic ectodermal dysplasia, Rothmund-Thompson syndrome, Cockayne syndrome, Bloom syndrome, but the clinical spectrum of abnormalities and the evolution with age were not compatible with one of these diagnoses. Parental consanguinity is in favour of autosomal recessive inheritance.     

Evaluation of cause of deaths' validity using outcome measures from a prospective, population based cohort study in Tehran, Iran

Objective

The aim of this study was to evaluate the validity of cause of death stated in death certificates in Tehran using outcome measures of the Tehran Lipid and Glucose Study (TLGS), an ongoing prospective cohort study.

Methods

The cohort was established in 1999 in a population of 15005 people, 3 years old and over, living in Tehran; 3551 individuals were added to this population three years later. As part of cohort''s outcome measures, deaths occurring in the cohort are investigated by a panel of medical specialists (Cohort Outcome Panel-COP) and underlying cause of death is determined for each death. The cause of death assigned in a deceased''s original death certificate was evaluated against the cause of death determined by COP and sensitivity and positive predictive values (PPV) were determined. In addition, determinants of assigning accurate underlying cause of death were determined using logistic regression model.

Result

A total of 231 death certificates were evaluated. The original death certificates over reported deaths due to neoplasms and underreported death due to circulatory system and transport accidents. Neoplasms with sensitivity of 0.91 and PPV of 0.71 were the most valid category. The disease of circulatory system showed moderate degree of validity with sensitivity of 0.67 and PPV of 0.78. The result of logistic regression indicated if the death certificate is issued by a general practitioner, there is 2.3 (95% CI 1.1, 5.1) times chance of being misclassified compared with when it is issued by a specialist. If the deceased is more than 60 years, the chance of misclassification would be 2.5 times (95% CI of 1.1, 5.9) compared with when the deceased is less than 60 years.     

Relationship between Auditory and Cognitive Abilities in Older Adults

Objective

The objective was to evaluate the association of peripheral and central hearing abilities with cognitive function in older adults.

Methods

Recruited from epidemiological studies of aging and cognition at the Rush Alzheimer’s Disease Center, participants were a community-dwelling cohort of older adults (range 63–98 years) without diagnosis of dementia. The cohort contained roughly equal numbers of Black (n=61) and White (n=63) subjects with groups similar in terms of age, gender, and years of education. Auditory abilities were measured with pure-tone audiometry, speech-in-noise perception, and discrimination thresholds for both static and dynamic spectral patterns. Cognitive performance was evaluated with a 12-test battery assessing episodic, semantic, and working memory, perceptual speed, and visuospatial abilities.

Results

Among the auditory measures, only the static and dynamic spectral-pattern discrimination thresholds were associated with cognitive performance in a regression model that included the demographic covariates race, age, gender, and years of education. Subsequent analysis indicated substantial shared variance among the covariates race and both measures of spectral-pattern discrimination in accounting for cognitive performance. Among cognitive measures, working memory and visuospatial abilities showed the strongest interrelationship to spectral-pattern discrimination performance.

Conclusions

For a cohort of older adults without diagnosis of dementia, neither hearing thresholds nor speech-in-noise ability showed significant association with a summary measure of global cognition. In contrast, the two auditory metrics of spectral-pattern discrimination ability significantly contributed to a regression model prediction of cognitive performance, demonstrating association of central auditory ability to cognitive status using auditory metrics that avoided the confounding effect of speech materials.     

Predicting Cognitive Function from Clinical Measures of Physical Function and Health Status in Older Adults

Introduction

Current research suggests that the neuropathology of dementia—including brain changes leading to memory impairment and cognitive decline—is evident years before the onset of this disease. Older adults with cognitive decline have reduced functional independence and quality of life, and are at greater risk for developing dementia. Therefore, identifying biomarkers that can be easily assessed within the clinical setting and predict cognitive decline is important. Early recognition of cognitive decline could promote timely implementation of preventive strategies.

Methods

We included 89 community-dwelling adults aged 70 years and older in our study, and collected 32 measures of physical function, health status and cognitive function at baseline. We utilized an L1–L2 regularized regression model (elastic net) to identify which of the 32 baseline measures were strongly predictive of cognitive function after one year. We built three linear regression models: 1) based on baseline cognitive function, 2) based on variables consistently selected in every cross-validation loop, and 3) a full model based on all the 32 variables. Each of these models was carefully tested with nested cross-validation.

Results

Our model with the six variables consistently selected in every cross-validation loop had a mean squared prediction error of 7.47. This number was smaller than that of the full model (115.33) and the model with baseline cognitive function (7.98). Our model explained 47% of the variance in cognitive function after one year.

Discussion

We built a parsimonious model based on a selected set of six physical function and health status measures strongly predictive of cognitive function after one year. In addition to reducing the complexity of the model without changing the model significantly, our model with the top variables improved the mean prediction error and R-squared. These six physical function and health status measures can be easily implemented in a clinical setting.     

Prognosis of iron chlorosis in pear (Pyrus communis L.) and peach (Prunus persica L. Batsch) trees using bud, flower and leaf mineral concentrations

Background and Aims

The possibility of using tree materials in early phenological stages, such as dormant buds and flowers, for the prognosis of Fe deficiency occurring later in the year has been studied in peach and pear trees.

Methods

Thirty-two peach trees and thirty pear trees with different Fe chlorosis degrees were sampled in different commercial orchards. In peach, samples included flower buds, vegetative buds, bud wood, flowers and leaves at 60 and 120?days after full bloom (DAFB). In pear, samples included buds, bud wood, flowers and leaves at 60 and 120?days DAFB. Leaf chlorophyll was assessed (SPAD) at 60 and 120 DAFB. Sampling was repeated for 3–5?years depending on the materials. Mineral nutrients measured were N, P, K, Ca, Mg, Fe, Mn, Zn and Cu.

Results

The relationships between the nutrient concentrations in the different materials and leaf SPAD were assessed using four different statistical approaches: i) comparison of means depending on the chlorosis level, ii) correlation analysis, iii) principal component analysis, and iv) stepwise multiple regression. In all cases, significant associations between nutrients and SPAD were found. The best-fit multiple regression curves obtained for the multi-year data set provided good prediction in individual years.

Conclusions

Results found indicate that it is possible to carry out the prognosis of Fe chlorosis using early materials such as buds and flowers. The relationships obtained were different from those obtained in previous studies using a single orchard. The different methods of analysis used provided complementary data.