Mel Bartley | University College London (original) (raw)

Papers by Mel Bartley

Journal of Epidemiology and Community Health, Oct 1, 2000

International Journal of Epidemiology, Aug 1, 2002

International Journal of Epidemiology, Jul 28, 2004

Journal of Epidemiology and Community Health, Feb 1, 2005

The present discussion paper sets forward a model within the life course perspective of how the s... more The present discussion paper sets forward a model within the life course perspective of how the social becomes biological. The model is intended to provide a framework for thinking about such questions as how does social class get into the molecules, cells and tissues of the body to produce social class differences in life expectancy and cause of death? A categorisation of social exposures and biological processes is suggested; and some principles governing their interrelations proposed. The paper ends by suggesting two public health applications of this approach.

Revue d'Épidémiologie et de Santé Publique, 2014

toute tentative de modélisation basée sur des algorithmes non paramétriques sera associée à un ta... more toute tentative de modélisation basée sur des algorithmes non paramétriques sera associée à un taux de convergence plus lent que la modèle paramétrique adapté. Enfin, il existe à ce jour de nombreux modèles non paramétriques, de telle sorte que la décision de choisir l'un plus que l'autre remet l'utilisateur dans la situation de faire une hypothèse sur la meilleure façon de modéliser ses données. Le principe du Super Learner est de ne faire aucune hypothèse en fournissant une libraire la plus riche possible en modèles paramétriques et non paramétriques. L'algorithme détermine ensuite à partir des données en utilisant la validation croisée et une fonction de perte appropriée quel est, au sein de la librairie, le modèle le plus adapté pour modéliser les données. Les modèles sont classés en fonction de leur performance et un nouvel algorithme hybride et orignal est construit sous la forme d'une combinaison pondérée des meilleurs modèles inclus dans la librairie. Les performances théoriques et empiriques du Super Learner ont fait l'objet de validations préalables. Nous montrons dans le présent travail que l'utilisation du Super Learner pour diverses questions pratiques épidémiologie clinique est associée à des résultats intéressants. Nous montrons par exemple l'intérêt d'une telle approche pour modéliser la mortalité en réanimation ou pour estimer le score de propension dans le cadre de l'estimation d'un paramètre causal en présence de données observationnelle. Mot clé Modélisation Déclaration d'intérêts L'auteur n'a pas transmis de déclaration de conflits d'intérêts.

Sociology of Health & Illness, 2008

Journal of the Royal Statistical Society. Series A (Statistics in Society), 1996

Summary measures of morbidity are necessary for research and health policy purposes but those use... more Summary measures of morbidity are necessary for research and health policy purposes but those used at present may provide unreliable estimates of the magnitude of social class differences in health. Class differences in illness behaviour will affect self-reported health ...

The European Journal of Public Health, 1994

Page 1. EUROPEAN OURNAL OF PUBLIC HEALTH 1994; 4:131-144 REVIEW Explanations for socio-economic d... more Page 1. EUROPEAN OURNAL OF PUBLIC HEALTH 1994; 4:131-144 REVIEW Explanations for socio-economic differentials in mortality Evidence from Britain and elsewhere GEORGE DAVEY SMITH, DAVID BLANE, MEL BARTLEY * ...

Is it just community development revamped?

International Journal for Equity in Health, 2020

BackgroundSocial class is frequently used as a means of ranking the population to expose inequali... more BackgroundSocial class is frequently used as a means of ranking the population to expose inequalities in health, but less often as a means of understanding the social processes of causation. We explored how effectively different social class mechanisms could be measured by longitudinal cohort data and whether those measures were able to explain health outcomes.MethodsUsing a theoretically informed approach, we sought to map variables within the National Child Development Study (NCDS) to five different social class mechanisms: social background and early life circumstances; habitus and distinction; exploitation and domination; location within market relations; and power relations. Associations between the SF-36 physical, emotional and general health outcomes at age 50 years and the social class measures within NCDS were then assessed through separate multiple linear regression models. R2values were used to quantify the proportion of variance in outcomes explained by the independent v...

Social Science & Medicine, 2019

A large body of literature has shown marked differences in the average levels of resources and ch... more A large body of literature has shown marked differences in the average levels of resources and child well-being across different family structures. Studies have examined cognitive, educational and behavioural outcomes; less is known about differentials in physical health, and about dynamics in early childhood. Furthermore, up to the present time, less emphasis has been placed on describing the underlying mechanisms relating childhood experience of family structure to health. In this paper, we hypothesize that socioeconomic characteristics and family structure trajectories will affect everyday , more proximal processes (material, behavioral and family stress pathways) directly experienced by the child, which will in turn affect child health. Using the UK Millennium Cohort Study, a nationally representative cohort of over 19,000 children born in 2001 and living in the UK shortly thereafter, we employ Graphical Chain Models to map the processes linking family structure trajectories to three physical health outcomes at age 5: overweight/obesity, respiratory health, and accidental injury. We construct family trajectories to highlight two components: status (distinguishing between married, cohabiting and single parents), and (in)stability. We show that both status, the (in)stability of that status, and their interplay, are important components of family structure trajectories which correlate to children's early physical health. Analyses highlight the relative importance of distinct pathways across different health outcomes. As well as some outcome-specific paths, we find that "family stress" variables appeared to underscore the relationship between family structure and child physical health, pointing to the importance of such variables in understanding how family structure relates to early child health.

SSM - Population Health, 2018

Social Science & Medicine, 1999

Many studies have attempted to understand observed social variations in cardiovascular disease in... more Many studies have attempted to understand observed social variations in cardiovascular disease in terms of sets of intermediate or confounding risk factors. Tests of these models have tended to produce inconsistent evidence. This paper examines the relationships to cardiovascular risk factors or two theoretically based measures of social position. It shows that the strength of the relationships between social position and cardiovascular risk factors varies according to the definition of social position which is used: there is a closer relationship between most health behaviours and the Cambridge scale, an indicator of &amp;amp;amp;#39;general social advantage and lifestyle&amp;amp;amp;#39;, whereas the Erikson-Goldthorpe schema, which is based on employment relations and conditions, is more strongly related to work control and breathlessness. The implications of these findings for understanding the conflicting evidence in other studies of health inequalities are then discussed. The paper concludes that inconsistencies between studies may be in part due to unexamined differences between the conceptual bases of the measures of social position they use, combined with a failure to make explicit the hypothetical mechanisms of effect. If neither the conceptual basis of the measure of social position, nor the links between social position and health outcome tested in each study are clear, inconsistencies between studies will be difficult to interpret, making policy recommendations highly problematic.

Scientists never conducted research in isolation from the outside world, but learn edinstead how ... more Scientists never conducted research in isolation from the outside world, but learn edinstead how to 'do impact'. Mel Bartley argues that case studies of research that has had impact on the wider world shows that we need to make an attempt to understand our research influence, and time to critically reflect on this process.

Longitudinal and Life Course Studies, 2016

In an opening paper Delpierre et al. explore the concept of allostatic load. The impact of the en... more In an opening paper Delpierre et al. explore the concept of allostatic load. The impact of the environment on our biological systems is summarised by the concept of embodiment. The biological embedding of social conditions could therefore be a relevant mechanism to partly explain the social gradient in health. A key issue is how to measure the 'physiological reality'the biological expression of embodiment at individual and population levels. Allostatic load (AL) has been proposed as a measure of the overall cost of adapting to the environment and may be a relevant tool or concept for measuring the way we have embodied our environment. Social inequalities in health may be partly explained by the embodiment of social environments, and AL may allow us to measure and compare embodiment between socioeconomic groups. However, before operationalising AL, a number of issues deserve further exploration. Among these, the choice of biological systems, and variables within each system, that should be included to remain 'loyal' to the theory of biological multisystem wastage underlying AL and the most appropriate methodological approach to be used to build an AL score, are particularly important. Moreover, studies analysing the link between adverse environments (physical, chemical, nutritional, psychosocial) across the life course and AL remain rare. Such studies require cohorts with data on socioeconomic and psychosocial environments over the life course, with multiple biological measures, made at various stages across the life span. The development and maintenance of these cohorts is essential to continue exploring the promising results that could enhance our understanding of the genesis of the social gradient in health by measuring embodiment.