Jorge G. Morel

Efficacy of an encapsulated probiotic Bifidobacterium infantis 35624 in women with irritable bowel syndrome.

BACKGROUND:Probiotic bacteria exhibit a variety of properties, including immunomodulatory activity, which are unique to a particular strain. Thus, not all species will necessarily have the same therapeutic potential in a particular condition. We have preliminary evidence that Bifidobacterium infantis 35624 may have utility in irritable bowel syndrome (IBS).OBJECTIVES:This study was designed to confirm the efficacy of the probiotic bacteria B. infantis 35624 in a large-scale, multicenter, clinical trial of women with IBS. A second objective of the study was to determine the optimal dosage of probiotic for administration in an encapsulated formulation.METHODS:After a 2-wk baseline, 362 primary care IBS patients, with any bowel habit subtype, were randomized to either placebo or freeze-dried, encapsulated B. infantis at a dose of 1 × 106, 1 × 108, or 1 × 1010, cfu/mL for 4 wk. IBS symptoms were monitored daily and scored on to a 6-point Likert scale with the primary outcome variable being abdominal pain or discomfort. A composite symptom score, the subjects global assessment of IBS symptom relief, and measures of quality of life (using the IBS-QOL instrument) were also recorded.RESULTS:B. infantis 35624 at a dose of 1 × 108 cfu was significantly superior to placebo and all other bifidobacterium doses for the primary efficacy variable of abdominal pain as well as the composite score and scores for bloating, bowel dysfunction, incomplete evacuation, straining, and the passage of gas at the end of the 4-wk study. The improvement in global symptom assessment exceeded placebo by more than 20% (p < 0.02). Two other doses of probiotic (1 × 106 and 1 × 1010) were not significantly different from placebo; of these, the 1 × 1010 dose was associated with significant formulation problems. No significant adverse events were recorded.CONCLUSIONS:B. infantis 35624 is a probiotic that specifically relieves many of the symptoms of IBS. At a dosage level of 1 × 108 cfu, it can be delivered by a capsule making it stable, convenient to administer, and amenable to widespread use. The lack of benefits observed with the other dosage levels of the probiotic highlight the need for clinical data in the final dosage form and dose of probiotic before these products should be used in practice.

Large Cluster Results for Two Parametric Multinomial Extra Variation Models

Abstract Two parametric extra variation models are considered. Approximate closed-form expressions are given for the Fisher information matrices. The expressions are useful in computing maximum likelihood estimates and obtaining large cluster efficiencies. A simulation study shows that the approximations perform very well even in clusters of moderate size. The models are applied in illustrative examples. A goodness-of-fit test is developed that is applicable even when the cluster sizes are unequal. The null distribution of the test statistic is shown to be well approximated by a chi-squared distribution. For the cluster size configurations in the examples, the test also has high power in distinguishing between the two models considered. The goodness-of-fit test shows that the new model provides adequate description of the data from the three experiments designed to study induced mutagenic effect.

An improved method for the computation of maximum likeliood estimates for multinomial overdispersion models

In this article, we consider the maximum likelihood estimation of two commonly used overdispersion models, namely, the Dirichlet-multinomial distribution (DM), due to Mosimann (Biometrika 49 (1962) 65), and a finite mixture distribution (FM) proposed by Morel and Nagaraj (Biometrika 80 (1993) 363), and Neerchal and Morel (J. Amer. Statist. Assoc. 93 (1998) 1078). These models have been successfully used in the literature for modeling overdispersion in multinomial data. Maximum likelihood estimation of the parameters of these models using the classical Fisher scoring method poses certain computational challenges. In the case of DM, the challenges are overcome by noting that the Fisher information matrix can be computed using the beta-binomial distribution (BB), which is the univariate version of DM. On the other hand, in the case of FM, an approximation theorem can be used to obtain a two-stage procedure for computing the maximum likelihood estimates. Simulation results show that the two-stage procedure is faster without loosing any accuracy.

Clustered binary logistic regression in teratology data using a finite mixture distribution

The beta-binomial distribution introduced by Skellam has been applied in many teratology problems for modelling the litter effect. Recently, Morel and Nagaraj proposed a new distribution for modelling cluster multinomial data when the clustering is believed to be caused by clumped sampling. It turns out that the distribution is a mixture of two binomial distributions and accommodates the estimation of an additional parameter to account for intra-litter effect. The new distribution arises from a cluster mechanism in which some individuals within a cluster exhibit the same behaviour while the remaining individuals from the cluster react independently of each other. Such a mechanism is a natural model in teratology problems, where typically a genetic trait is passed with a certain probability to the foetuses of the same litter. In this article, we use the new distribution to model binary responses with logistic regression. We analyse data from a teratology experiment to demonstrate that the new model provides a useful addition to current methodology. The experiment investigates the synergistic effect of the anticonvulsant phenytoin and trichloropopene oxide on the prenatal development of inbred mice. In a simulation study we investigate the type I error rate and the power of the maximum likelihood ratio test when the data follow a finite mixture distribution.

Toxic Shock Syndrome: Characterization of Human Immune Responses to TSST-1 and Evidence for Sensitivity Thresholds

Noninvasive vaginal infections by Staphylococcus aureus strains producing the superantigen TSST-1 can cause menstrual toxic shock syndrome (mTSS). With the objective of exploring the basis for differential susceptibility to mTSS, the relative responsiveness to TSST-1 of healthy women has been investigated. Peripheral blood mononuclear cells from healthy donors were incubated with purified TSST-1 or with the T-cell mitogen phytohemmaglutinin (PHA), and proliferation was measured. The concentrations of TSST-1 and PHA required to elicit a response equivalent to 15% of the maximal achievable response (EC15) were determined. Although with PHA, EC15 values were comparable between donors, subjects could be classified as being of high, medium, or low sensitivity based on responsiveness to TSST-1. Sensitivity to TSST-1-induced proliferation was associated with increased production of the cytokines interleukin-2 and interferon-γ. When the entire T lymphocyte population was considered, there were no differences between sensitivity groups with respect to the frequency of cells known to be responsive to TSST-1 (those bearing CD3(+) Vβ2(+)). However, there was an association between sensitivity to TSST-1 and certain HLA-class II haplotypes. Thus, the frequencies of DR7DQ2, DR14DQ5, DR4DQ8, and DR8DQ4 haplotypes were greater among those with high sensitivity, a finding confirmed by analysis of responses to immortalized homozygous B cell lines. Collectively, the results reveal that factors other than neutralizing antibody and the frequency of Vβ2(+) T lymphocytes determine immunological responsiveness to TSST-1. Differential responsiveness of lymphocytes to TSST-1 may form the basis of interindividual variations in susceptibility to mTSS.

Maximum-likelihood estimation of the random-clumped multinomial model as a prototype problem for large-scale statistical computing

Numerical methods are needed to obtain maximum-likelihood estimates (MLEs) in many problems. Computation time can be an issue for some likelihoods even with modern computing power. We consider one such problem where the assumed model is a random-clumped multinomial distribution. We compute MLEs for this model in parallel using the Toolkit for Advanced Optimization software library. The computations are performed on a distributed-memory cluster with low latency interconnect. We demonstrate that for larger problems, scaling the number of processes improves wall clock time significantly. An illustrative example shows how parallel MLE computation can be useful in a large data analysis. Our experience with a direct numerical approach indicates that more substantial gains may be obtained by making use of the specific structure of the random-clumped model.

Sample size determination for alternate periods of use study designs with binary responses.

In this article, we consider several study designs that arise in practice, which are variations of standard crossover designs. Often, they may result from modifications made to a standard crossover design due to practical considerations. Characteristic features of the studies we are concerned with are (a) treatments consist of external use of products with little or no possibility of carry over effects, and (b) the periods of use are dictated by the subjects or by some specific event, such as diaper leakage or menstrual flow. We consider a number of such study designs for estimating the difference in the efficacy of two treatments or test products. We provide brief descriptions of studies to motivate the study design, the underlying data structure, and computations of the variances of the usual unbiased estimators of the difference in efficacy, and the sample size formulas. The situations considered here cover a number of popular crossover designs. The objective of our work is to provide guidance to members of a wide audience on how to answer the sample size question for their own nonstandard situations. We conclude the article with a brief report on a simulation study we conducted to investigate the impact of estimation on the sample size determination and consequently on the actual power realized in an effort to promote the “best practice” of checking whether the recommended sample sizes indeed achieve the desired level of power.

A covariance matrix that accounts for different degrees of extraneous variation in multinomial responses

A covariance matrix structure that generalizes the single scale covariance matrix usually used in multinomial responses with extraneous variation is presented. The proposed covariance matrix allows for various levels of extraneous variation, and might be useful in modeling either extra variation or under dispersion. An explicit representation of the proposed covariance matrix, as well as a meaningful interpretation in terms of regression ideas, are provided. An example is presented for illustration