Comparison of methods for auto-coding causation of injury narratives.

Manually reading free-text narratives in large databases to identify the cause of an injury can be very time consuming and recently, there has been much work in automating this process. In particular, the variations of the naïve Bayes model have been used to successfully auto-code free text narratives describing the event/exposure leading to the injury of a workers' compensation claim. This paper compares the naïve Bayes model with an alternative logistic model and found that this new model outperformed the naïve Bayesian model. Further modest improvements were found through the addition of sequences of keywords in the models as opposed to consideration of only single keywords. The programs and weights used in this paper are available upon request to researchers without a training set wishing to automatically assign event codes to large data-sets of text narratives. The utility of sharing this program was tested on an outside set of injury narratives provided by the Bureau of Labor Statistics with promising results.

Auxological perspectives on 'growth' in DOHaD.

David Barker established growth as a seminal link between early development and later health attainment and disease risk. This was nothing less than a paradigm shift in health and medicine, turning the focus of disease causality away from contemporary environmental influences to earliest growth as a time when functional anatomy and physiology sets in place critical structures and function for a lifetime. Barker's prodigious work investigated time- and place-specific interactions between maternal condition and exogenous environmental influences, focusing on how growth unfolds across development to function as a mechanistic link to ensuing health. Subsequent applications do not always attend to the specificity and sensitivity issues included in his original work, and commonly overlook the long-standing methods and knowledge base of auxology. Methodological areas in need of refinement include enhanced precision in how growth is represented and assessed. For example, multiple variables have been used as a referent for 'growth,' which is problematic because different body dimensions grow by different biological clocks with unique functional physiologies. In addition, categorical clinical variables obscure the spectrum of variability in growth experienced at the individual level. Finally, size alone is a limited measure as it does not capture how individuals change across age, or actually grow. The ground-breaking notion that prenatal influences are important for future health gave rise to robust interest in studying the fetus. Identifying the many pathways by which size is realized permits targeted interventions addressing meaningful mechanistic links between growth and disease risk to promote health across the lifespan.

Trajectories of body mass index amongst children who develop type 2 diabetes as adults.

OBJECTIVE: Type 2 diabetes (T2D) is a heterogeneous disorder. The aim of this study was to examine the trajectories of childhood growth associated with T2D. DESIGN AND SUBJECTS: A total of 13 345 individuals born in Helsinki, Finland between 1934 and 1944 were included in the study. The participants' growth had been recorded in detail during childhood, and 11.7% (n = 1558) had been diagnosed with T2D. We divided the cohort around the median body mass index (BMI) at 11 years. Body composition and glucose tolerance were assessed in a clinical subsample (n = 2003) in adulthood. RESULTS: Two pathways of growth were associated with T2D. Both began with low weight and BMI at birth. In one, persistent low BMI through infancy was followed by a rapid increase in BMI in childhood. Amongst individuals with a BMI at 11 years above the median value, the odds ratio for T2D associated with a one z-score increase in BMI between 2 and 11 years was 1.31 (95% confidence interval 1.21-1.42, P < 0.001). In the other pathway, low BMI at birth, accompanied by short length at birth, was followed by low BMI in childhood. Most women who developed diabetes followed this trajectory; they developed T2D at a lower BMI and lower fat percentage than women with a BMI above the median at 11 years of age. CONCLUSIONS: Two pathways of early growth trigger T2D. Low fat deposition leading to thinness at birth and during infancy results in fat acquisition during childhood. Reduced linear growth leading to short length at birth is associated with lower body fat percentage in adulthood but increased risk of developing diabetes.

Small head circumference at birth and early age at adiposity rebound.

AIMS: The adiposity rebound is the age in childhood when body mass index is at a minimum before increasing again. The age at rebound is highly variable. An early age is associated with increased obesity in later childhood and adult life. We have reported that an early rebound is predicted by low weight gain between birth and 1 year of age and resulting low body mass index at 1 year. Here, we examine whether age at adiposity rebound is determined by influences during infancy or is a consequence of foetal growth. Our hypothesis was that measurements of body size at birth are related to age at adiposity rebound. METHODS: Longitudinal study of 2877 children born in Helsinki, Finland, during 1934-1944. RESULTS: Early age at adiposity rebound was associated with small head circumference and biparietal diameter at birth, but not with other measurements of body size at birth. The mean age at adiposity rebound rose from 5.8 years in babies with a head circumference of ≤33 cm to 6.2 in babies with a head circumference of >36 cm (P for trend = 0.007). The association between thinness in infancy and early rebound became apparent at 6 months of age. It was not associated with adverse living conditions. In a simultaneous regression, small head circumference at birth, high mother's body mass index and tall maternal stature each had statistically significant trends with early adiposity rebound (P = 0.002, <0.001, 0.004). CONCLUSION: We hypothesize that the small head size at birth that preceded an early adiposity rebound was the result of inability to sustain a rapid intra-uterine growth trajectory initiated in association with large maternal body size. This was followed by catch-up growth in infancy, and we hypothesize that this depleted the infant's fat stores.

Development and evaluation of a Naïve Bayesian model for coding causation of workers' compensation claims.

INTRODUCTION: Tracking and trending rates of injuries and illnesses classified as musculoskeletal disorders caused by ergonomic risk factors such as overexertion and repetitive motion (MSDs) and slips, trips, or falls (STFs) in different industry sectors is of high interest to many researchers. Unfortunately, identifying the cause of injuries and illnesses in large datasets such as workers' compensation systems often requires reading and coding the free form accident text narrative for potentially millions of records. METHOD: To alleviate the need for manual coding, this paper describes and evaluates a computer auto-coding algorithm that demonstrated the ability to code millions of claims quickly and accurately by learning from a set of previously manually coded claims. CONCLUSIONS: The auto-coding program was able to code claims as a musculoskeletal disorders, STF or other with approximately 90% accuracy. IMPACT ON INDUSTRY: The program developed and discussed in this paper provides an accurate and efficient method for identifying the causation of workers' compensation claims as a STF or MSD in a large database based on the unstructured text narrative and resulting injury diagnoses. The program coded thousands of claims in minutes. The method described in this paper can be used by researchers and practitioners to relieve the manual burden of reading and identifying the causation of claims as a STF or MSD. Furthermore, the method can be easily generalized to code/classify other unstructured text narratives.

Resource allocation in utero and health in later life.

The way that a fetus obtains and allocates nutritional resources has profound consequences for its life-long health. Under the new developmental model for the origins of chronic disease, the causes to be identified are linked to normal variations in the processes of feto-placental development, that are associated with differences in the supply of nutrients to the baby. These differences programme the function of a few key systems that are linked to chronic disease, including the immune system, anti-oxidant defences, inflammation, and the number and quality of stem cells. There is not a separate cause for each different disease. Which chronic disease originates during development may depend more on timing than on qualitative differences in experience.

The use of regularity as estimated by approximate entropy to distinguish saltatory growth.

A nonlinear dynamics metric, approximate entropy (ApEn), is investigated as a diagnostic method for distinguishing between mathematical models, and the underlying mechanistic hypotheses that purport to describe the same time series experimental observations. ApEn measures the occurrence of pattern regularity within a time series, and is used here to investigate growth patterns in daily length growth. The notion investigated is that ApEn distributions for competing time series patterns expressed as mathematical formulations can be modelled by Monte Carlo and bootstrap methods and compared to the ApEn values for an original experimental data series. If the ApEn values for the different models do not overlap, then it is expected that ApEn can be utilized to distinguish these models and hypotheses, and to provide statistical assessment for the underlying biological patterns in experimental data. The conclusion is that the ApEn metric is successful as a time series diagnostic tool. It is a model-independent statistic that clearly differentiates saltatory growth from slowly varying continuous models of growth and serves to further document the saltatory nature of growth. This is a unique application of approximate entropy, illustrating the broad applicability of ApEn to biological time series, with the specific example of discriminating a saltatory growth process in longitudinal growth data. Future investigations of regularity in longitudinal time series in human biology with ApEn statistics are suggested.

Mixed distribution analysis identifies saltation and stasis growth.

A maximum likelihood method of mixed distribution analysis is investigated for its utility as a method for the identification of saltation and stasis in longitudinal growth data. Daily infant growth data that have been previously identified to follow a saltatory growth process are employed. This is a novel application of the finite mixed distribution analysis (MDA), a method designed to objectively identify the presence of one or more Gaussian populations. The null hypothesis is that a single Gaussian distribution best describes the incremental growth data. This would be compatible with smooth, slowly varying daily growth patterns. This study explores whether or not two distinctive populations are evident in incremental saltatory growth data, as postulated by the saltation and stasis observations. The analysis is important in providing a growth model-independent test for the presence of saltation and stasis by a separate statistical assessment with none of the saltatory algorithm assumptions. The finite mixed distribution analysis identifies that each individual's incremental growth data is statistically best described as a mixture consisting of two components, or two populations of increments (chi-square, p < 0.05). For each individual, one of these populations is centred about a zero increment, and is compatible with the previous evidence of stasis intervals. The second population of data points is characterized by unique distributions for each individual, compatible with the previous observation that infants grow by unique patterns of growth saltations in both amplitude and frequency. The percentage of data points that fall within each of the two unique finite mixture distributions (FMDs) is similar to the proportions of discrete saltation and stasis intervals previously identified by the saltation and stasis method. Thus, the FMD analysis lends support to the nature of growth as a saltatory process characterized by two states in the daily growth of these infants. By contrast with the saltatory algorithm, which is applied to the original serial growth measurements, the mixed distribution analysis employs increments removed from their time relationships. The lack of time series sequence information precludes the mixed distribution method from reconstructing specific temporal patterns of saltatory growth. The present analysis reiterates that individual growth patterns are statistically unique and cannot be reconstructed or identified from group data.

Child factor in measurement dependability.

A primary consideration in longitudinal growth studies is the identification of growth from error components. While previous research has considered matters of measurement accuracy and reproducibility in detail, few reports have investigated the errors of measurement due to aspects of the physiology and cooperation of the child. The present study directly assesses this source of measurement undependability for the first time. Investigation of total measurement error variance in 925 recumbent length replicates taken over stasis intervals in growth identifies that between 60% and 70% of total measurement unreliability is due to a child factor undependability. Individual differences are significant and longitudinal growth analyses should consider two to three times the technical error of measurement statistic as a reasonable estimate of the total unreliability for any single measurement of an infant's recumbent length. These results raise issues regarding analytic methods as applied to serial growth data.

A novel cold-sensitive allele of the rate-limiting enzyme of fatty acid synthesis, acetyl coenzyme A carboxylase, affects the morphology of the yeast vacuole through acylation of Vac8p.

The yeast vacuole functions both as a degradative organelle and as a storage depot for small molecules and ions. Vacuoles are dynamic reticular structures that appear to alternately fuse and fragment as a function of growth stage and environment. Vac8p, an armadillo repeat-containing protein, has previously been shown to function both in vacuolar inheritance and in protein targeting from the cytoplasm to the vacuole. Both myristoylation and palmitoylation of Vac8p are required for its efficient localization to the vacuolar membrane (Y.-X. Wang, N. L. Catlett, and L. S. Weisman, J. Cell Biol. 140:1063-1074, 1998). We report that mutants with conditional defects in the rate-limiting enzyme of fatty acid synthesis, acetyl coenzyme A carboxylase (ACC1), display unusually multilobed vacuoles, similar to those observed in vac8 mutant cells. This vacuolar phenotype of acc1 mutant cells was shown biochemically to be accompanied by a reduced acylation of Vac8p which was alleviated by fatty acid supplementation. Consistent with the proposed defect of acc1 mutant cells in acylation of Vac8p, vacuolar membrane localization of Vac8p was impaired upon shifting acc1 mutant cells to nonpermissive condition. The function of Vac8p in protein targeting, on the other hand, was not affected under these conditions. These observations link fatty acid synthesis and availability to direct morphological alterations of an organellar membrane.

Electrospray ionization tandem mass spectrometry (ESI-MS/MS) analysis of the lipid molecular species composition of yeast subcellular membranes reveals acyl chain-based sorting/remodeling of distinct molecular species en route to the plasma membrane.

Nano-electrospray ionization tandem mass spectrometry (nano-ESI-MS/MS) was employed to determine qualitative differences in the lipid molecular species composition of a comprehensive set of organellar membranes, isolated from a single culture of Saccharomyces cerevisiae cells. Remarkable differences in the acyl chain composition of biosynthetically related phospholipid classes were observed. Acyl chain saturation was lowest in phosphatidylcholine (15.4%) and phosphatidylethanolamine (PE; 16.2%), followed by phosphatidylserine (PS; 29.4%), and highest in phosphatidylinositol (53.1%). The lipid molecular species profiles of the various membranes were generally similar, with a deviation from a calculated average profile of approximately +/- 20%. Nevertheless, clear distinctions between the molecular species profiles of different membranes were observed, suggesting that lipid sorting mechanisms are operating at the level of individual molecular species to maintain the specific lipid composition of a given membrane. Most notably, the plasma membrane is enriched in saturated species of PS and PE. The nature of the sorting mechanism that determines the lipid composition of the plasma membrane was investigated further. The accumulation of monounsaturated species of PS at the expense of diunsaturated species in the plasma membrane of wild-type cells was reversed in elo3Delta mutant cells, which synthesize C24 fatty acid-substituted sphingolipids instead of the normal C26 fatty acid-substituted species. This observation suggests that acyl chain-based sorting and/or remodeling mechanisms are operating to maintain the specific lipid molecular species composition of the yeast plasma membrane.

The Saccharomyces cerevisiae hyperrecombination mutant hpr1Delta is synthetically lethal with two conditional alleles of the acetyl coenzyme A carboxylase gene and causes a defect in nuclear export of polyadenylated RNA.

In a screen for mutants that display synthetic lethal interaction with hpr1Delta, a hyperrecombination mutant of Saccharomyces cerevisiae, we have isolated a novel cold-sensitive allele of the acetyl coenzyme A (CoA) carboxylase gene, acc1(cs), encoding the rate-limiting enzyme of fatty acid synthesis. The synthetic lethal phenotype of the acc1(cs) hpr1Delta double mutant was only partially complemented by exogenous fatty acids. hpr1Delta was also synthetically lethal with a previously isolated, temperature-sensitive allele of ACC1, mtr7 (mRNA transport), indicating that the lethality of the acc1(cs) hpr1Delta double mutant was not allele specific. The basis for the interaction between conditional acc1 alleles and hpr1Delta was investigated in more detail. In the hpr1Delta mutant background, acetyl-CoA carboxylase enzyme activity was reduced about 15-fold and steady-state levels of biotinylated Acc1p and ACC1 mRNA were reduced 2-fold. The reduced Acc1p activity in hpr1Delta cells, however, did not result in an altered lipid or fatty acid composition of the mutant membranes but rendered cells hypersensitive to soraphen A, an inhibitor of Acc1p. Similar to mtr7, hpr1Delta and acc1(cs) mutant cells displayed a defect in nuclear export of polyadenylated RNA. Oversized transcripts were detected in hpr1Delta, and rRNA processing was disturbed, but pre-mRNA splicing appeared wild type. Surprisingly, the transport defect of hpr1Delta and acc1(cs) mutant cells was accompanied by an altered ring-shaped structure of the nucleolus. These observations suggest that the basis for the synthetic lethal interaction between hpr1Delta and acc1 may lie in a functional overlap of the two mutations in nuclear poly(A)+ RNA production and export that results in an altered structure of the nucleolus.

Wrinkles induced by the use of smoothing procedures applied to serial growth data.

This paper elucidates the effects of moving average filters when applied to serial growth measurements. This is a question of interest because smoothing procedures are inherently part of a number of analytical methods presently employed in auxological analyses. Particular attention is paid to sequential growth data analysed to identify what has been described as pulsatile, saltation and stasis patterns or mini-growth spurts. When applied to pulsatile, or saltatory, time series data the process of smoothing itself creates artifactual temporal patterns in the time series data similar to previously described mini growth spurts while removing the actual pulsatile characteristics of the data. These observations illustrate that smoothing approaches add noise to time series data while removing meaningful patterns in the original data sequence. Analyses employing such approaches produce results that include waveforms or other fluctuations compatible with an underlying pulsatile driving mechanism, but do not necessarily reflect the temporal characteristics of the original biological process.

An example of variation and pattern in saltation and stasis growth dynamics.

The serial data from two siblings, aged 6.6 and 7.5 years of age at the initiation of the study, measured each evening for total standing height during 365 days, are analysed by two methods to investigate the nature of the underlying growth pattern. The saltation and stasis model, designed to identify the presence of statistically significant pulses in sequential data, is compared for goodness-of-fit to first to sixth degree polynomial functions, used to investigate the presence of a slowly varying smooth continuous function in the data, and high order polynomials of the same degree of flexibility as the individual's saltation and stasis results. The saltation and stasis model is found to better-fit the experimental data than the slowly varying smooth continuous functions (p < 0.01 to 0.001). The timing characteristics of the saltation and stasis patterns are investigated and the temporal patterns are suggestive of a non-random, aperiodical deterministic system.

Is growth saltatory? The usefulness and limitations of frequency distributions in analyzing pulsatile data.

Several investigators have proposed that descriptive statistics can be employed to identify and discriminate growth patterns. These studies assumed that the shape of the frequency distribution of daily growth velocities (FDGVs) is diagnostic in differentiating between a pattern of growth characterized by smooth, continuous daily acquisition and a pattern of growth characterized by a discontinuous, i.e. pulsatile process. The FDGV from a saltation and stasis, i.e. episodic or pulsatile, growth pattern was assumed to be bimodal or significantly skewed to the right, whereas a continuous growth function was assumed to be approximately Gaussian. The use of FDGV characteristics is an unprecedented approach to the analysis of longitudinal growth data and was not previously validated for this use. The present study investigates the performance characteristics of the FDGV method by Monte-Carlo simulations of known saltatory, i.e. pulsatile, growth patterns. These analyses show that the FDGV for a saltation and stasis growth process can be either unimodal or bimodal and either skewed to the right or to the left. Data collection frequency, measurement error, and total study duration all determine the shape of the FDGV and the statistical significance of the results. If the FDGV is highly skewed, then it is consistent with saltatory growth. However, if the FDGV is not highly skewed, then it is consistent with both the saltatory model and a smooth, continuous growth model, and thus, the results are ambiguous. We conclude that FDGV analysis is not a valid method to exclude saltation and stasis growth processes in longitudinal growth studies.

Problems in the aging of skeletal juveniles: perspectives from maturation assessments of living children.

We employ samples of children of known chronological age to demonstrate the significance of random and systematic effects on maturation in both dental and skeletal development. Differences between chronological age for dental age in young healthy Canadian children can be as much as 100% of the actual age of the children. For skeletal development by reference to Greulich-Pyle standards, three samples of known-age children from Mexico document parallel effects: 1) 183 six-year-old children have skeletal-based ages with a 95% confidence interval of 4-8 years; 2) 80% of 217 4.0-4.5-year-old children are underaged by 1-3 years; and 3) 130 children of skeletal age between 39 and 44 months are actually between 4 and 7.4 chronological years of age. The Mexican samples are drawn from a population living under conditions of environmental stress with chronic mild to moderate protein-energy malnutrition and moderate to high levels of infectious disease. These children may parallel those from the past, whose remains are studied by skeletal biologists or paleoanthropologists. Our findings reinforce concerns expressed in extant studies regarding the accuracy of age-at-death reconstructions.