Biomedical subjects

Michael G Kenward

Publications and source records attributed to Michael G Kenward.

10 recordsLinked to original sources

PubMed2006

Social capital and mental health: a comparative analysis of four low income countries.

Women and the poor are disproportionately affected by common mental disorders (CMD), and women in low income countries are particularly at risk. Social capital may explain some of the geographical variation in CMD, but the association between social capital and CMD in low income countries has rarely been studied. This paper aims to explore the relationship between individual and ecological measures of social capital and maternal CMD in four low income countries. Cross-sectional data from the Young Lives (YL) study with information across 234 communities in Peru, Ethiopia, Vietnam and Andhra Pradesh (India) were used. The mental health of mothers of one-year-old children (n=6909), and the individual cognitive and structural social capital of all respondents was assessed. Ecological social capital was calculated by aggregating individual responses to the community level. Multi-level modelling was used to explore the association between individual and ecological (community level) social capital and maternal CMD in each of the four countries, adjusting for a wide range of individual and community level confounders. The analysis shows that individual cognitive social capital is associated with reduced odds of CMD across all four countries. The results for structural social capital are more mixed and culturally specific, with some aspects associated with increased odds of CMD. This suggests that structural social capital has context-specific effects and cognitive social capital more universal effects on maternal CMD.

Adolescent↗

PubMed2006

The analysis of longitudinal data using mixed model L-splines.

L-splines are a large family of smoothing splines defined in terms of a linear differential operator. This article develops L-splines within the context of linear mixed models and uses the resulting mixed model L-spline to analyze longitudinal data from a grassland experiment. In the spirit of time-series analysis, a periodic mixed model L-spline is developed, which partitions data into a smooth periodic component plus smooth long-term trend.

Biomass↗

PubMed2005

Maternal consumption of dairy products during pregnancy and lactation, and the development of cow's milk antibodies in the offspring.

OBJECTIVE: To assess whether the maternal consumption of milk and milk products affects development of cow's milk (CM) antibodies in infants. DESIGN: A randomized pilot trial using food frequency questionnaires (mothers) and food records (infants). SETTING: Families with a newborn infant with increased HLA-DQB1-conferred risk of type 1 diabetes and at least one first-degree relative affected by type 1 diabetes from 16 hospitals in Finland between April 1995 and November 1997. SUBJECTS AND INTERVENTION: Infants randomized to receive a hydrolysed formula when breast milk was not available during their first 6-8 mo (n=112). Of these, 13 dropped out by the age of 3 mo and two were excluded due to incomplete CM antibody data. RESULTS: Maternal milk protein intake from cheese during pregnancy was inversely related to IgA-class antibody titres to beta-lactoglobulin (BLG) and casein (CAS) at 3 mo, and to IgA antibody titres to BLG at 6 mo. Maternal consumption of raw milk products during lactation was positively related to the development of IgA antibody titres to CAS at 6 mo, and inversely correlated to IgG antibody titres to bovine serum albumin (BSA) and IgA antibody titres to CAS at 2 y. Maternal cheese consumption was inversely related to the IgG antibody titres to CM formula and CAS and to the IgA antibody titres to CAS in early infancy. CONCLUSIONS: Few associations were established between maternal CM protein intake and CM protein antibody levels in the infants. The milk and milk products taken by the mother differed in their impact on the emerging CM antibody response in the offspring.

Animals↗

PubMed2005

Direct likelihood analysis versus simple forms of imputation for missing data in randomized clinical trials.

BACKGROUND: In many clinical trials, data are collected longitudinally over time. In such studies, missingness, in particular dropout, is an often encountered phenomenon. METHODS: We discuss commonly used but often problematic methods such as complete case analysis and last observation carried forward and contrast them with broadly valid and easy to implement direct-likelihood methods. We comment on alternatives such as multiple imputation and the expectation-maximization algorithm. RESULTS: We apply these methods in particular to data from a study with continuous outcomes. The outcomes are modelled using a general linear mixed-effects model. The bias with CC and LOCF is established in the case study and the advantages of the direct-likelihood approach shown. CONCLUSIONS: We have established formal but easy to understand arguments for a shift towards a direct-likelihood paradigm when analysing incomplete data from longitudinal clinical trials, necessitating neither imputation nor deletion.

Data Collection↗

PubMed2005

Serum alpha-tocopherol concentrations and risk of type 1 diabetes mellitus: a cohort study in siblings of affected children.

Animal models have indicated that alpha-tocopherol may protect against type 1 diabetes mellitus (DM1). Epidemiological data on the subject are scarce. The objective of this study was to evaluate the association of serum alpha-tocopherol concentration and risk of DM1 in a cohort of initially non-diabetic siblings of children affected by DM1 (n = 722). We used two study designs: 1) Siblings who progressed to DM1 were compared with control siblings who remained negative for DM1-associated autoantibodies in a nested case-control study design. 2) All siblings with DM1-associated autoantibodies were prospectively followed for DM1. In both designs, high concentrations of serum alpha-tocopherol tended to be associated with a lower risk of DM1 (p = 0.08 and 0.09, respectively). Although the results did not reach statistical significance, they support the hypothesis that high alpha-tocopherol levels may protect against DM1.

Adolescent↗

PubMed2004

The analysis of repeated 'direct' measures of change illustrated with an application in longitudinal imaging.

The use of repeated measures of an outcome variable to improve statistical power and precision in randomized clinical trials and cohort studies is well documented. Linear mixed models have great utility in the analysis of such studies in many medical applications including imaging. However, in imaging studies and other applications the basic outcome can be a 'direct' measure of change in a variable, as opposed to a difference calculated by subtraction of one measured value from another. The correlation structure of such repeated measures of 'direct' change, in particular the non-independence of within-person consecutive measures, adds complexity to the analysis. In this paper, we present a family of hierarchical mixed models for the analysis of such data and explain how to implement them using standard statistical software. We illustrate the use of our models with data from a cohort of patients with Alzheimer's disease.

Alzheimer Disease↗

PubMed2004

Blood biochemistry and the risk of cancer.

A longitudinal study based on a serum sample bank was carried out in Finland to find out the association between biochemical substances and the subsequent risk of cancer. The objective was to evaluate the consistency between means of individually estimated levels of these compounds and levels based on pooling. Levels of alpha-tocopherol, beta-carotene, retinol, retinol-binding protein, and ceruloplasmin were estimated by primary site and sex and partly by age and morphology. The concentrations in pooled samples were consistently lower than the averages of the individual samples. On the basis of individual samples, all the five biochemical compounds had a rather consistent protective effect on the risk of cancers at most primary sites. This protective effect disappeared in the pool analyses, and more than half of exposure contrasts showed an opposite sign. For ceruloplasmin, the effect of pooling was smaller but not negligible. The results of this study emphasize the demand to standardize the collecting, handling, and analysing of samples in serum banks. They are, furthermore, consistent with the hypothesis that pooling of biochemical samples affects the levels of the substances and may affect the conclusions of epidemiological studies on causes of diseases.

Biomarkers, Tumor↗

PubMed2004

Analyzing incomplete longitudinal clinical trial data.

Using standard missing data taxonomy, due to Rubin and co-workers, and simple algebraic derivations, it is argued that some simple but commonly used methods to handle incomplete longitudinal clinical trial data, such as complete case analyses and methods based on last observation carried forward, require restrictive assumptions and stand on a weaker theoretical foundation than likelihood-based methods developed under the missing at random (MAR) framework. Given the availability of flexible software for analyzing longitudinal sequences of unequal length, implementation of likelihood-based MAR analyses is not limited by computational considerations. While such analyses are valid under the comparatively weak assumption of MAR, the possibility of data missing not at random (MNAR) is difficult to rule out. It is argued, however, that MNAR analyses are, themselves, surrounded with problems and therefore, rather than ignoring MNAR analyses altogether or blindly shifting to them, their optimal place is within sensitivity analysis. The concepts developed here are illustrated using data from three clinical trials, where it is shown that the analysis method may have an impact on the conclusions of the study.

Antidepressive Agents↗

PubMed2004

Validation of surrogate markers in multiple randomized clinical trials with repeated measurements: canonical correlation approach.

Part of the recent literature on the evaluation of biomarkers as surrogate endpoints starts from a multitrial context, which leads to a definition of validity in terms of the quality of both trial-level and individual-level association between the surrogate and true endpoints (Buyse et al., 2000, Biostatistics1, 49-67). These authors concentrated on cross-sectional continuous responses. However, in many randomized clinical studies, repeated measurements are encountered on either or both endpoints. A challenge in this setting is the formulation of a simple and meaningful concept of "surrogacy."Alonso et al. (2003, Biometrical Journal45, 931-945) proposed the variance reduction factor (VRF) to evaluate surrogacy at the individual level. They also showed how and when this concept should be extended to study surrogacy at the trial level. Here, we approach the problem from the natural canonical correlation perspective. We define a class of canonical correlation functions that can be used to study surrogacy at the trial and individual level. We show that the VRF and the R2 measure defined by Buyse et al. (2000) follow as special cases. Simulations are conducted to evaluate the performance of different members of this family. The methodology is illustrated on data from a meta-analysis of five clinical trials comparing antipsychotic agents for the treatment of chronic schizophrenia.

Antipsychotic Agents↗

PubMed2003

Clinical diagnosis of Achilles tendinopathy with tendinosis.

OBJECTIVE: To evaluate sensitivity, specificity, reproducibility, and predictive value of palpation of the painful arc sign and of the Royal London Hospital test in 10 patients with Achilles tendinopathy and in 14 asymptomatic subjects. DESIGN: Test-retest study. SETTING: University teaching hospital. PARTICIPANTS: Ten male athletes on the waiting list for exploration of one of their Achilles tendons for tendinopathy of the main body of the tendon attended a special clinic. Each was invited to bring at least one athlete of the same sex in the same discipline aged within 2 years of themselves with no history and no symptoms of Achilles tendinopathy. A total of 14 controls were thus recruited. MAIN OUTCOME MEASURES: Pain and tenderness following performance of palpation, the painful arc sign, and the Royal London Hospital test. RESULTS: There were no statistically significant differences at the 5% level among the effects of investigator or between morning and afternoon measurements for any of the three measurement methods. There was no evidence of a difference of the three assessment methods (p > 0.05). When the three methods were combined, the overall sensitivity was 0.586 (confidence interval [CI], 0.469-0.741), and the overall specificity was 0.833 (CI, 0.758-0.889). CONCLUSIONS: In patients with tendinopathy of the Achilles tendon with a tender area of intratendinous swelling that moves with the tendon and whose tenderness significantly decreases or disappears when the tendon is put under tension, a clinical diagnosis of tendinopathy can be formulated, with a high positive predictive chance that the tendon will show ultrasonographic and histologic features of tendinopathy.

Achilles Tendon↗