Insights on bias and information in group-level studies

Ecological and aggregate data studies are examples of group-level studies. Even though the link between the predictors and outcomes is not preserved in these studies, inference about individual-level exposure effects is often a goal. The disconnection between the level of inference and the level of analysis expands the array of potential biases that can invalidate the inference from group-level studies. While several sources of bias, specifically due to measurement error and confounding, may be more complex in group-level studies, two sources of bias, cross-level and model specification bias, are a direct consequence of the disconnection. With the goal of aligning inference from individual versus group-level studies, I discuss the interplay between exposure and study design. I specify the additional assumptions necessary for valid inference, specifically that the between- and within-group exposure effects are equal. Then cross-level inference is possible. However, all the information in the group-level analysis comes from between-group comparisons. Models where the group-level analysis provides even a small percentage of information about the within-group exposure effect are most susceptible to model specification bias. Model specification bias can be even more serious when the group-level model isn't derived from an individual-level model.     

Insights from stable light isotopes on enamel defects and weaning in Pliocene herbivores

A high prevalence of enamel hypoplasia in several herbivores from the early Pliocene Langebaanweg locality, South Africa, indicates general systemic stress during the growing years of life. The presence of several linear enamel hypoplasias per tooth crown in many teeth further suggest that these stress events may be episodic. The δ¹⁸O values along tooth crowns of mandibular second molars ofSivatherium hendeyi (Artiodactyla, Giraffidae) were used to investigate the cause of the stress events in this tooth type. Results show that weaning in this fossil giraffid occurred at a similar ontogenetic age to that in extant giraffes, and that the observed enamel hypoplasia towards the base of this tooth type manifested post-weaning. Further, high-resolution oxygen isotope analyses acrossS. hendeyi third molars suggest that the entire development of defective tooth crowns occurred under conditions of increased aridity in which the cool, rainy part of the seasonal cycle was missing. The high prevalence of this defect in many herbivores suggests that climatic conditions were not favourable. This study reiterates the value of stable isotope analyses in determining both the behaviour of fossil animals and the environmental conditions that prevailed during tooth development.     

Microstructural characteristics of thin biofilms through optical and scanning electron microscopy

The combination of a conventional optical microscope with a specially designed glass flow cell was used to visualize in situ biofilms formed on opaque thin biomaterials through a simple non-invasive way (optical microscopy of thin biofilms, OMTB). Comparisons of OMTB with scanning electron microscopy (SEM) images were made. Thin metallic dental biomaterials were used as substrata. They were immersed in a synthetic saliva and in a modified Mitis–Salivarius medium inoculated with a consortium of oral microorganisms. To study the effect of bacterial motility, Pseudomonas fluorescens cultures were also used. The processes which give rise to the formation of the biofilm were monitored through OMTB. Biofilm microstructures like pores, water channels, streamers and chains of Streptococci, attached to the surface or floating in the viscous interfacial environment, could be distinguished. Thickness and roughness of the biofilms formed on thin substrata could also be evaluated. Distortions introduced by pretreatments carried out to prepare biological materials for SEM observations could be detected by comparing OMTB and SEM images. SEM images (obtained at high magnification but ex situ, not in real time and with pretreatment of the samples) and OMTB images (obtained in situ, without pretreatments, in real time but at low magnification) in combination provided complementary information to study biofilm processes on thin substrata.     

Small-sample bias and corrections for conditional maximum-likelihood odds-ratio estimators

A number of small-sample corrections have been proposed for the conditional maximum-likelihood estimator of the odds ratio for matched pairs with a dichotomous exposure. I here contrast the rationale and performance of several corrections, specifically those that generalize easily to multiple conditional logistic regression. These corrections or Bayesian analyses with informative priors may serve as diagnostics for small-sample problems. Points are illustrated with a small exact performance comparison and with an example from a study of electrical wiring and childhood leukemia. The former comparison suggests that small-sample bias may be more prevalent than commonly realized.     

The Evolution of Enamel Microstructure: How Important Is Amelogenin?

The shape, size, and orientation of enamel prisms have heretofore been thought to be controlled solely by the shape of the Tomes' process. It is known, however, that amelogenin proteins play an important role in enamel deposition and maturation and it is possible that they contribute independently to enamel structure. Using a phylogenetic framework, we clarify the role of amelogenin proteins in the formation of enamel microstructure. We found a negative association between evolutionary changes in amelogenin protein sequences and enamel complexity: amelogenin evolution slows as enamel complexity increases. This is probably because selective constraints on amelogenin increase as enamel complexity increases. Monotremes, which have lost their adult dentition, have particularly high rates of amelogenin evolution while rodents, which have very complex enamel, have very low rates. There is a positive correlation between the number of different amelogenin proteins in a given species and the complexity of its enamel microstructure. An increased number of amelogenins may be necessary for the formation of multiple enamel types in the same tooth. Alternative splicing of amelogenin exons, which allows multiple protein products to be produced from the same gene, may be a key innovation in the diversification of enamel microstructure.     

Accelerated Evolution of Cytochrome b in Simian Primates: Adaptive Evolution in Concert with Other Mitochondrial Proteins?

We have sequenced the cytochrome b gene of Horsfield's tarsier, Tarsius bancanus, to complete a data set of sequences for this gene from representatives of each primate infraorder. These primate cytochrome b sequences were combined with those from representatives of three other mammalian orders (cat, whale, and rat) in an analysis of relative evolutionary rates. The nonsynonymous nucleotide substitution rate of the cytochrome b gene has increased approximately twofold along lineages leading to simian primates compared to that of the tarsier and other primate and nonprimate mammalian species. However, the rate of transversional substitutions at fourfold degenerate sites has remained uniform among all lineages. This increase in the evolutionary rate of cytochrome b is similar in character and magnitude to that described previously for the cytochrome c oxidase subunit II gene. We propose that the evolutionary rate increase observed for cytochrome b and cytochrome c oxidase subunit II may underlie an episode of coadaptive evolution of these two proteins in the mitochondria of simian primates. Received: 15 December 1997 / Accepted: 24 February 1998     

Temperature and pH-Dependent Supramolecular Self-Assembly of Amelogenin Molecules: A Dynamic Light-Scattering Analysis

Evidence for the molecular self-assembly of amelogenin proteins to form quasi-spherical particles (“nanospheres”) in solution, bothin vitroandin vivo,has recently been documented. A particle-size distribution analysis of dynamic light-scattering data was undertaken to investigate the influence of temperature on this molecular self-assembly process at three different pH's. The long-term objective was to correlate these observations to the unusual physiochemical characteristics of the protein, to improve understanding of the molecular mechanisms involved in the generation of amelogenin “nanospheres” and understanding of their putative relation to amelogenin functionin vivo. We analyzed data using two different algorithms: Dynamics and DynaLS. It was found that at pH 8, in a temperature range between 5 and 25°C, the size of the recombinant amelogenin nanospheres is monodisperse, giving rise to particles of 15–18 nm in hydrodynamic radius. However, heterogeneous distribution of particle size was observed at temperature ranges between 27 and 35°C, becoming monodisperse again with larger particles (60–70 nm) after the temperature was elevated to 37–40°C. We interpret these results to suggest that amelogenin molecular self-association possesses a second stage assembly process at temperatures of 30–35°C, creating larger entities which apparently are structured and stable at 37–40°C. The effect of pH on the size of amelogenin “aggregates” was much more noticeable at 37°C compared to that at 25°C. This observation suggests that at physiological temperature (i.e., 37°C) amelogenin molecular self-assembly is extremely sensitive to pH changes. This finding supports the notion that local pH changes in the microenvironment of the enamel extracellular matrix may play critical roles in controlling the structural organization of the organic matrix framework.     

Citric acid production and morphology of Aspergillus niger as functions of the mixing intensity in a stirred tank and a tubular loop bioreactor

The relationship between Aspergillus niger morphology and citric acid production was investigated in two reactor systems with different configurations, a tubular loop and a stirred tank bioreactor, with operating volumes of 6 and 8 dm³, respectively. Morphology was quantified by image analysis. In each system, morphology, characterized by the parameter P (mean convex perimeter of clumps), and citric acid production, were agitation-dependent and closely linked. Increased agitation caused a reduction of clump sizes and results when both reactors demonstrate that the parameter P should not exceed a threshold value in order to achieve increased productivities. The results obtained from the two reactors were in agreement, both qualitatively and quantitatively. Reducing the fundamentally different mixing conditions of the two bioreactors to the order of the dimensionless mixing parameter relative mixing time (τ_m), results showed that the loop simulated the stirred tank. Also, relationships valid for one system accurately described the results obtained from the other system, demonstrating the validity of the relationship between morphology and productivity for the particular fermentation, regardless of the reactor type. Previous attempts to evaluate the use of loop configurations as scale-up tools and their performance as bioreactors, neglected the morphology of the producer micro-organisms. This study demonstrated the close link between morphology and productivity for citric acid production by A. niger, and identified a morphology parameter that was used successfully to characterize the process performance.     

Strategies of leaf expansion in Ficus carica under semiarid conditions

Leaf area expansion, thickness and inclination, gas exchange parameters and relative chlorophyll content were analysed in field‐grown fig (Ficus carica L.) leaves over time, from emergence until after full leaf expansion (FLE). Ficus carica leaves showed a subtle change in shape during the early stages of development, and FLE was reached within ca. 30 days after emergence. Changes in leaf thickness and inclination after FLE demonstrated good adaptation to environmental conditions during summer in areas with a Mediterranean climate. Changes in gas exchange parameters and relative chlorophyll content showed that F. carica is a delayed‐greening species, reaching maximum values 20 days after FLE. Correlation analysis of datasets collected during leaf expansion, confirmed dependence among structural and functional traits in F. carica. Pn was directly correlated with stomatal conductance (Gs), transpiration (E), leaf area (LA) and relative chlorophyll content up to FLE. The effect of pruning on leaf expansion, a cultural technique commonly applied in this fruit tree, was also evaluated. Although leaf development in pruned branches gave a significantly higher relative leaf area growth rate (RGR_l) and higher LA than non‐pruned branches, no significant differences were found in other morphological and physiological traits, indicating no pruning effect on leaf development. All studied morphological and physiological characteristics indicate that F. carica is well adapted to semiarid conditions. The delayed greening strategy of this species is discussed.     

A Global View on Parametric and Nonparametric Approaches to the Analysis of Ordered Categorical Data

Rank approaches are very common in the analysis of ordered categorical data but can only be interpreted on an experiment‐wise level. Therefore, parametric tests from linear models, although based on metric structures, are used frequently to analyze this type of data. So the questions arise 1. what parametric tests measure in this context and 2. whether the rank approach could be modified to achieve a global level of interpretation. A possible solution to question 2. offers the so called ridit approach, which is based on known reference distributions. In this paper we discuss a global view that shows how rank analysis and ridit analyses are related and how parametric procedures fit into the same framework. The use of the uniform distribution as a reference in the ridit approach gives an explanation to question 1. The asymptotic multivariate normality of the effect estimators is shown and robust test statistics are discussed. Type I and type II error rates are examined in simulation studies and the approach is applied to a toxicological example. (© 2004 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)