Search

La vie en couple est-elle bénéfique pour les hommes et pour les femmes, ou pour les hommes au détriment des femmes ? Comment un événement biographique comme la perte du conjoint s’insère-t-il dans le processus du vieillissement ? Nous abordons ces questions à travers une pratique quotidienne qui se déroule largement dans l’espace domestique tout en étant un bon indicateur d’un style de vie conforme aux recommandations de santé et aux goûts dominants : la consommation quotidienne de légumes. Nous utilisons la cohorte épidémiologique Gazel de l’Inserm dont les 20 625 enquêtés sont suivis depuis 1989. Les hommes voient leur consommation de légumes décroitre plus que les femmes quand ils connaissent une rupture d’union. Leur consommation est aussi plus sensible à la position sociale de leur conjointe. La consommation de légumes des femmes ne diminue qu’après la rupture d’une union avec un homme de situation socioprofessionnelle modeste. Nous concluons que, dans notre population d’étude vieillissante, la conjugalité bénéficie aux deux conjoints, mais plus aux hommes qu’aux femmes. Cet article propose un apport méthodologique sur le traitement des non-réponses dans des données de cohorte, et théorique en discutant la possibilité d’une intersection entre genre, classe et statut conjugal dans la sphère domestique.

AIM To derive from a validated questionnaire a periodontal screening score (PESS), intended as a user-friendly tool to identify individuals at risk of periodontitis in epidemiological studies. METHODS A French 12-item self-reported questionnaire was developed by translating English questions previously used for periodontitis screening and surveillance. After a cognitive evaluation, the questionnaire was validated in a sample of 232 individuals (mean age: 46.1 ± 12.6 years) receiving full-mouth periodontal examination, including probing pocket depth and clinical attachment level recordings. Case definition was based on the American Academy of Periodontology/Centers for Disease Control and Prevention criteria. Logistic regression analyses and C-statistics were used to assess the validity and accuracy of the questionnaire and to develop the PESS. RESULTS The sample was constituted of 109 individuals with severe periodontitis, who were compared with 123 individuals with no/moderate periodontitis. The questionnaire had moderate-to-high accuracy in identifying severe cases; the PESS (calculated on five self-report items, age, and smoking) showed a sensitivity of 78.9% and a specificity of 74.8%, with an area under the receiver operating characteristics curve of 0.821. CONCLUSION The PESS represents a valuable and accurate tool to screen for severe periodontitis at the population level.

Hostility has been found to be positively associated with alcohol intake in cross-sectional studies. Our aim was to examine prospectively the long-lasting association of hostility with alcohol consumption.We included 10,612 men and 3834 women from the French Gazel cohort with mean ages in 1993 of 48.6 (SD=2.9) and 45.7(SD=4.2), respectively. Hostility (i.e., total, cognitive and behavioral) was assessed in 1993 with the Buss and Durkee Hostility Inventory. Alcohol consumption was self-reported annually from 1994 to 2014. Hostility scores were introduced successively in general linear mixed models with annual alcohol consumption in drinks per week as dependent variable. Multivariable analyses were adjusted for age, occupational status, marital status, retirement status and depression score. All the analyses were stratified by sex.Among men (women), 83.0% (76.2%) completed at least 75% of all annual assessment of alcohol consumption over a 21-year follow-up. In univariate analysis, alcohol consumption was associated with total and behavioral hostility in both sex (all p0.001). In multivariable analyses, these associations remained significant with a greater size effect for behavioral hostility. Estimated means of alcohol consumptions ranged from 10.50 [95CI%: 10.01-10.92] drinks per week to 13.32 [95%CI: 12.90-13.74] in men and from 4.09 [95%CI: 3.71-4.46] to 5.78 [95%CI: 5.39-6.17] in women, for the first and last quartiles respectively (p trends0.001 and all pairwise comparisons0.01). Similar effects were observed among participants with at-risk alcohol consumption at baseline.In both men and women, behavioral hostility predicted alcohol consumption over a 21-year follow-up. Interventions aiming at modulating behavioral hostility may help reducing its long-lasting influence on alcohol consumption.

OBJECTIVE To measure the impact of diabetes on work cessation, i.e., on the risks of work disability, early retirement, and death while in the labor force. RESEARCH DESIGN AND METHODS We used data from the GAZEL prospective cohort of 20,625 employees of the French national gas and electricity company “EDF-GDF.” We identified 506 employees with diabetes and randomly selected 2,530 nondiabetic employed control subjects matched for major sociodemographic and occupational characteristics. Using a multistate Cox model, we estimated hazard ratios (HRs) comparing the risks of transition from employment to disability, retirement, and death over time between participants with versus without diabetes. RESULTS Employment rate decreased more rapidly in participants with diabetes (51.9 and 10.1% at 55 and 60 years, respectively) compared with nondiabetic participants (66.5 and 13.4%, respectively). Participants with diabetes had significantly increased risks of transition from employment to disability (HR 1.7 [95% CI 1.0–2.9]), retirement (HR 1.6 [1.5–1.8]), and death (HR 7.3 [3.6–14.6]) compared with participants without diabetes. Between 35 and 60 years, each participant with diabetes lost an estimated mean time of 1.1 year in the workforce (95% CI 0.99–1.14) compared with a nondiabetic participant. CONCLUSIONS Our results provide evidence for a profound negative impact of diabetes on workforce participation in France. Social and economic consequences are major for patients, employers, and society—a burden that is likely to increase as diabetes becomes more and more common in the working-aged population.

AimsTo examine the association of alcohol consumption over 10 years with cognitive performance in different socio-economic groups.DesignProspective cohort study, the French GAZEL study.SettingFrance.ParticipantsEmployees of France's national electricity and gas company.MeasurementsAlcohol intake was assessed annually, beginning in 1992, using questions on frequency and quantity of alcoholic beverages consumed in a week; used to define mean consumption and trajectory of alcohol intake over 10 years. Cognitive performance among participants aged >= 55 years (n = 4073) was assessed in 2002-04 using the Digit Symbol Substitution Test (DSST), a measure of psychomotor speed, attention and reasoning. Occupational position at age 35 and education were used as the markers of socio-economic position.FindingsAll analyses were stratified by socio-economic position. In the low occupational group, participants consuming a mean of more than 21 drinks per week had 2.1 points lower (95% CI: -3.9, -0.3) DSST score compared to those consuming four to 14 drinks per week. In participants with primary school education, the corresponding difference was 3.6 points (95% CI: -7.1, -0.0). No association between alcohol consumption and cognitive performance was observed in the intermediate and high socio-economic groups, defined using either occupation or education. Analysis of trajectories of alcohol consumption showed that in the low socio-economic groups large increase or decrease in alcohol consumption was associated with lower cognitive scores compared to stable consumption.ConclusionsOur results suggest that high alcohol consumption is associated with poorer cognitive performance only in the low socio-economic group, due possibly to greater cognitive reserve in the higher socio-economic groups.

International audience; This paper examines trends in obesity rates and education-related absolute and relative inequalities in obesity over the last 40 years in France. Data are drawn from the French Decennial Health Surveys of 1970, 1980, 1991 and 2003. The difference in obesity rates between the least- and most-educated, the Slope Index of Inequality, is used to estimate absolute inequalities in obesity. The ratio of the corresponding rates, the Relative Index of Inequality, reveals the relative inequalities in obesity. Obesity rates were similar in men and women, but educational inequalities were greater in women. Obesity rates were similar over the first three surveys but increased for all in the 2003 survey. This increase was accompanied by increases in absolute inequalities in men (P = 0.04) from a Slope Index of Inequality of 4.80% (95% confidence interval [CI] = 2.27, 7.32) to 8.64% (95% CI = 5.97, 11.32) and women (P = 0.004) from 8.90% (95% CI = 6.18, 11.63) to 14.57% (95% CI = 11.83, 17.32). Relative inequalities in obesity remained stable over the 40 years. Recent increase in obesity rates in France is accompanied by increases in absolute education-related inequalities, while relative inequalities have remained stable; this suggests that obesity rates have increased at a much faster rate in the low-education groups.

International audience; BACKGROUND: Analysis of change in health status using data from two waves can be examined either adjusted or unadjusted for baseline health status. The effect of socioeconomic position (SEP) on cognitive change was assessed using both these strategies and the implications of the analyses are discussed. METHODS: Data from 1261 men and 483 women of the Whitehall II cohort study, aged 50-55 years at wave 1, were used. Cognition was assessed at both waves using a test of verbal memory, and two tests of verbal fluency. Analysis of variance (ANOVA) was used to estimate the effect of SEP on change score and analysis of covariance (ANCOVA) was used to estimate this effect adjusted for the baseline cognitive score. The ANCOVA estimates were corrected for bias due to measurement error (estimated based on 3-month test-retest). Finally, ANCOVA estimates were examined for increasing levels of measurement error. RESULTS: The results of the ANOVA suggest no effect of SEP on cognitive decline. In contrast, the ANCOVA suggests significantly greater cognitive decline in the lower SEP groups. However, the ANCOVA estimates for the effect of wave 1 cognition show evidence for regression to the mean due to the presence of measurement error. The corrected ANCOVA estimates show no association between SEP and cognitive decline. CONCLUSIONS: Results from analysis of change using two waves of observational data, when adjusted for baseline, should be interpreted with caution.

International audience; To determine if self-rated health (SRH), a single-item measure of health status where individuals are asked to rate their own health, predicts mortality in a middle-aged sample and if the predictive ability of SRH diminishes with time. Data (6316 men and 3035 women) are drawn from the Whitehall II study. SRH and covariates were measured at baseline (1985-1988) when the average age of individuals was 44.5 years (SD = 6.1). The mortality follow-up was available for a mean of 17.5 years and was classified as having occurred in the first 10 years or the subsequent follow-up period (range 6 to 9 years). The association between SRH and mortality was assessed using a Cox regression model with relative index of inequality (RII) to summarize associations. There were no sex differences in the association between SRH and mortality in either the short (p = .39) or the long term (p = .16). Sex-adjusted short-term association (RII = 3.80; 95% confidence interval (CI) 2.28, 6.35) was significantly (p = .004) stronger than the long-term association (RII = 1.56; 95% CI 1.04, 2.34). Explanatory variables accounted for 80% of the SRH-mortality association in men and 29% in women. SRH predicts mortality equally well in men and women. However, the covariates explained a much larger proportion of the SRH-mortality relationship in men compared with women. In this middle-aged cohort, SRH predicts mortality strongly in the short term but only weakly in the long term.

Background The decline in participation rates in surveys, including epidemiological surveillance surveys, has become a real concern since it may increase nonresponse bias. The aim of this study is to estimate the contribution of a complementary survey among a subsample of nonrespondents, and the additional contribution of paradata in correcting for nonresponse bias in an occupational health surveillance survey. Methods In 2010, 10,000 workers were randomly selected and sent a postal questionnaire. Sociodemographic data were available for the whole sample. After data collection of the questionnaires, a complementary survey among a random subsample of 500 nonrespondents was performed using a questionnaire administered by an interviewer. Paradata were collected for the complete subsample of the complementary survey. Nonresponse bias in the initial sample and in the combined samples were assessed using variables from administrative databases available for the whole sample, not subject to differential measurement errors. Corrected prevalences by reweighting technique were estimated by first using the initial survey alone and then the initial and complementary surveys combined, under several assumptions regarding the missing data process. Results were compared by computing relative errors. Results The response rates of the initial and complementary surveys were 23.6% and 62.6%, respectively. For the initial and the combined surveys, the relative errors decreased after correction for nonresponse on sociodemographic variables. For the combined surveys without paradata, relative errors decreased compared with the initial survey. The contribution of the paradata was weak. Conclusion When a complex descriptive survey has a low response rate, a short complementary survey among nonrespondents with a protocol which aims to maximize the response rates, is useful. The contribution of sociodemographic variables in correcting for nonresponse bias is important whereas the additional contribution of paradata in correcting for nonresponse bias is questionable.

The aim of this study was to describe the change in vegetable consumption with ageing and the transition to retirement. Study subjects were the participants of the GAZEL prospective cohort (Gaz and Électricité de France) aged 40–49 years at inclusion in 1989 who retired between 1991 and 2008 (12 942 men and 2739 women). Four FFQ were completed from 1990 to 2009. We used multiple imputation by chained equations in order to avoid dropping incomplete cases. The OR for eating vegetables everyday was estimated as a function of ageing, retirement status and the place of lunch before retirement through generalised estimating equations. Analyses were stratified by sex, and models were adjusted for confounders, including current spousal status. In 1990, 17·7 % of men and 31 % of women reported eating vegetables daily. The odds of consuming vegetables everyday increased with ageing for both men and women. The usual place of lunch was home for less than half the sample before retirement and for almost every respondent after retirement. For those who changed their place of lunch, the association between being retired and the odds of eating vegetables daily was positive and significant. We found that, in this cohort, vegetable consumption increased with ageing. Retirement had an indirect effect on vegetable consumption mediated by changes in the place of lunch.

Background: Changes in life events may play a contributing role in changes in smoking behaviors. The objective was to examine the impact of stressful life events (SLEs) on smoking among French adults. Methods: We examined smoking prevalence in 20 625 employees of the French GAZEL cohort for up to 5 years before and after a SLE during three time periods (years −1 vs. −5; years +1 vs. −1; years +5 vs. +1). Repeated measures analysis of time series data indexed to events were used, employing generalized estimating equations. Results: For women, comparing 1 year after vs. 1 year before SLEs, decreased odds of smoking were found for employment promotion (OR: 0.80; 95% CI = 0.67–0.95), marriage (OR: 0.57; 95% CI = 0.48–0.68) and divorce (OR: 0.78; 95% CI = 0.68–0.90). Comparing 5 years after to 1 year after SLEs, women had decreased odds of smoking for important purchase (OR: 0.87; 95% CI = 0.79–0.96), children leaving home (OR: 0.83; 95% CI = 0.74–0.93), retirement (OR: 0.73; 95% CI = 0.64–0.83) and death of loved one (OR: 0.86; 95% CI = 0.79–0.93). For men, decreased odds of smoking were observed in all three time periods for all SLEs except when comparing 1 year before to 5 years before marriage (OR: 1.66; 95% CI = 1.09–2.52) and divorce (OR: 1.49; 95% CI = 1.25–1.77). Conclusion: Time surrounding SLEs during which individuals are susceptible to changing smoking behaviors may be an important consideration.

Objectives Job stress is a growing epidemiologic field in France. The aim of this study was to evaluate the levels of occupational psychosocial risk factors to which the French working population is exposed. It also focused on developing a job-exposure matrix. Methods This study used existing French national data on work conditions collected by the French Ministry of Labor in 1991, in which 20 929 workers were interviewed by questionnaire. The items of the questionnaire were retained that represented potential stressors. A principal component analysis was performed to summarize the data in terms of job-stress factors. Four independent variables (gender, age, occupation, and activity) were available, as well as scores for exposure to psychosocial risk factors. The CART (classification and regression tree) segmentation method was used to construct the job-exposure matrix. Results Fourteen psychosocial factors were identified and interpreted on the basis of the results of the principal component analysis and deepened by experts’ judgment. To take into account the well-known difference of distribution in occupations between men and women, a gender-specific matrix was developed. Conclusions This study is the first attempt to develop a job-exposure matrix in the area of job-stress factors in France. Thus the results have allowed the assessment of exposure to 14 psychosocial factors for all of the 455 categories of the French occupational classification for men and women separately.

Background: Job-exposure matrices (JEMs) applicable to the general population are usually constructed by using only the expertise of specialists. Aims: To construct a population based JEM for chemical agents from data based on a sample of French workers for surveillance purposes. Methods: The SUMEX job-exposure matrix was constructed from data collected via a cross-sectional survey of a sample of French workers representative of the main economic sectors through the SUMER-94 survey: 1205 occupational physicians questioned 48 156 workers, and inventoried exposure to 102 chemicals. The companies' economic activities and the workers' occupations were coded according to the official French nomenclatures. A segmentation method was used to construct job groups that were homogeneous for exposure prevalence to chemical agents. The matrix was constructed in two stages: consolidation of occupations according to exposure prevalence; and establishment of exposure indices based on individual data from all the subjects in the sample. Results: An agent specific matrix could be constructed for 80 of the chemicals. The quality of the classification obtained for each was variable: globally, the performance of the method was better for less specific and therefore more easy to assess agents, and for exposures specific to certain occupations. Conclusions: Software has been developed to enable the SUMEX matrix to be used by occupational physicians and other prevention professionals responsible for surveillance of the health of the workforce in France.

Contexte En sante comme dans d’autres domaines, les taux de reponse aux enquetes diminuent de plus en plus. La non-reponse est souvent problematique car elle peut entrainer des biais ; elle a egalement des consequences deleteres sur la variance des estimateurs. L’etude des biais de non-reponse n’est pas nouvelle en epidemiologie : plusieurs etudes ont deja montre que la participation a une enquete epidemiologique est liee a l’âge, a la categorie sociale, a l’etat de sante de la personne et aux comportements a risque pour la sante tels que la consommation d’alcool et de tabac. Nous nous placons ici dans le cadre d’enquetes de sante publique ou de surveillance epidemiologique ayant pour objectif de produire des estimations de prevalence extrapolables a une population d’interet. On distingue la non-reponse totale qui survient lorsque la personne enquetee ne repond a aucune question de l’enquete a de la non-reponse partielle, appelee plus communement donnee manquante, qui est rencontree si la personne enquetee repond a certaines questions de l’enquete, mais pas a toutes. Le biais de non-reponse depend de la variable d’interet, puisqu’il est fonction du produit de l’inverse de la probabilite de reponse moyenne et de la covariance entre la probabilite de reponse et la variable d’interet. Autrement dit, il y aura absence de biais si la probabilite de reponse est egale a 1 (tout le monde repond), ou si la covariance entre la probabilite de reponse et la variable d’interet est nulle. Les non-reponses peuvent etre classees selon trois types : non-reponse completement aleatoire (Missing Completely At Random ou MCAR), non-reponse aleatoire (Missing At Random ou MAR), non-reponse non aleatoire (Missing Not At Random ou MNAR). Dans le cas MCAR, il y a independance entre la probabilite de reponse et la variable d’interet. La prevalence estimee de la variable d’interet est sans biais. Dans le cas MAR, il y a independance entre la probabilite de reponse et la variable d’interet conditionnellement a d’autres variables notees X. Autrement dit, la probabilite de reponse et la variable d’interet partagent des causes communes X. Apres prise en compte de ces causes communes par des methodes appropriees, la prevalence estimee de la variable d’interet est sans biais. Quelle que soit la technique utilisee, la correction est possible a condition que l’ensemble des causes communes X soient disponibles pour l’ensemble des repondants et des non-repondants. Dans le cas MNAR, il n’y a pas independance entre la probabilite de reponse et la variable d’interet. Elle peut resulter soit d’un lien direct entre la variable d’interet et la probabilite de reponse, soit lorsqu’il n’a pas ete possible de prendre en compte l’ensemble des causes communes X. Dans ce cas, la prevalence estimee de la variable d’interet ne peut etre sans biais. Les methodes qui permettent de prendre en compte les causes communes X dans le cas MAR seront presentees : la reponderation (preferentiellement utilisee pour traiter la non-reponse totale) et l’imputation (preferentiellement utilisee pour traiter les donnees manquantes). Correction de la non-reponse totale par reponderation : l’exemple de la phase pilote de la cohorte Coset-RSI La non reponse totale traitee par reponderation est presentee ici et sera illustree a partir des donnees de la phase pilote de la cohorte Coset-RSI, cohorte pour la surveillance epidemiologique en lien avec le travail aupres d’actifs relevant du regime social des independants au moment de l’inclusion. La reponderation pour correction de la non-reponse consiste a augmenter les poids de sondage des repondants afin de compenser l’absence de reponse des non-repondants. La premiere etape est d’expliquer le mecanisme de non-reponse en le modelisant a partir des informations qui sont disponibles a la fois pour les repondants et pour les non-repondants. Cela suppose d’avoir de telles informations, dites variables auxiliaires, qui, par definition, ne sont pas issues du questionnaire. Parmi les variables auxiliaires disponibles, on selectionne les variables qui expliquent la non-reponse et qui sont liees aux variables d’interet evaluees par le questionnaire. La cohorte Coset-RSI a la particularite de disposer de nombreuses variables auxiliaires. En effet, outre les donnees de la base de sondage qui sont classiquement utilisees, les donnees de l’assurance maladie (SNIIRAM) sont renseignees pour tous les individus selectionnes pour etre enquetes. Pour etre utilisees, les donnees du SNIIRAM sont synthetisees sous la forme d’indicateurs disponibles a la fois pour les repondants et les non-repondants. La non-reponse est ensuite modelisee selon ces indicateurs via une regression logistique, qui permet d’obtenir pour chaque individu une probabilite de reponse predite. La deuxieme etape est la constitution de groupes de reponse homogene. Pour les constituer, la methode des scores est utilisee : les groupes correspondent aux quantiles de la distribution des probabilites de reponse predites. Dans chaque groupe ainsi constitue, on calcule ensuite le taux de reponse observe. L’inverse de ce taux de reponse est applique a tous les repondants du groupe, en multipliant leur poids de sondage initial par ce facteur correctif. Chaque etape de la methode de reponderation sera illustree a partir des donnees de la phase pilote de la cohorte Coset-RSI. L’impact de la correction de la non-reponse sera observe sur des prevalences mesurees dans l’enquete telles que le tabagisme, l’obesite ou la symptomatologie depressive. Traitement de la non-reponse partielle par imputation : l’exemple de la cohorte Constances La non-reponse partielle traitee par imputation est presentee ici et sera illustree a partir des donnees d’inclusion de la cohorte Constances, cohorte epidemiologique « generaliste » constituee a terme d’un echantillon de 200 000 adultes âges de 18 a 69 ans a l’inclusion et affilies au regime general de la securite sociale. L’imputation consiste a remplacer les valeurs manquantes par des valeurs plausibles. Les methodes d’imputation reposent sur les memes hypotheses que les methodes de ponderation, a savoir que les donnees sont manquantes au hasard, c’est-a-dire manquantes aleatoirement conditionnellement aux donnees observees. Les methodes d’imputation sont nombreuses : on distingue les methodes deterministes des methodes probabilistes et les imputations simples des imputations multiples. Nous presenterons deux methodes d’imputation qui peuvent etre utilisees si l’objectif est d’estimer la prevalence d’une variable d’interet et que toutes les variables auxiliaires sont non manquantes. La methode « hot deck par donneur » consiste a remplacer les valeurs manquantes d’un sujet « receveur » par celle d’un sujet « donneur » ne presentant pas de donnee manquante et ayant les memes caracteristiques (e.g. les memes variables auxiliaires) que celles du sujet « receveur ». L’imputation par modele de regression consiste a remplacer la valeur manquante d’un sujet par sa prediction issue d’un modele de regression expliquant la variable d’interet par les variables auxiliaires. Les prevalences sur donnees completes et sur donnees imputees seront comparees pour la symptomatologie depressive (mesuree par le CESD) et pour le questionnaire de Siegrist.

Objective: To explore the relation between risk factors (RF) and occupational mobility in working men. Setting: 20 000 volunteers working at the French National Electricity and Gas Company (GAZEL cohort). Participants: Men aged 43 to 53 years in 1992. Design: Three designs were used for analysis. (1) The association between occupational mobility experienced before 1992 and RF reported at that date was analysed among 10 383 men. (2) The predictive role of RF on occupational mobility over 1992–1999 was studied in a subsample of 4715 men. (3) Reciprocally, occupational mobility in 1985–1992 was analysed in relation to RF changes over 1993–1999. Main outcome measures: Self reported smoking status, excessive alcohol consumption, arterial hypertension, and overweight. Occupational mobility defined by any upward transition between senior executives and professionals/middle executives/employees, and workers. Results: (1) Cross sectionally, non-mobile men as their entry into the company had a higher risk of being smokers, excessive alcohol drinkers, and overweight in 1992 than mobile men. (2) Longitudinally, smokers and excessive alcohol drinkers in 1992 had a higher risk of non-mobility than, respectively, non-smokers and non-excessive alcohol drinkers. (3) Non-mobile men in 1985–1992 had a higher risk of becoming smokers, excessive alcohol drinkers, and hypertensive in 1993–1999 than upwardly mobile men. Conclusion: These results suggest a complex relation between RF and occupational mobility. A high level of RF, particularly health behaviours, might account for a selection process reducing upward occupational mobility. In turn, a lack of upward occupational mobility might be associated with an increased incidence of RF.

Objectives This study aimed at reconstructing changes in the frequency and levels of occupational asbestos exposure in France over the past century. Methods Work histories were collected during 11 population-based case-referent studies recently carried out in France, and an asbestos-specific job-exposure matrix including 10 625 jobs was used to estimate indices of past occupational asbestos exposure. The results were estimated from a sample of 4287 subjects, bootstrapped 200 times. Results The distribution of socioeconomic categories within the sample was compared with that of the general population in 1954, 1962, 1968, 1975, and 1982. The proportion of blue-collar workers was similar. The highest proportion of exposed subjects was found between 1950 and 1980. Around 10% of each 10-year age class was exposed to asbestos. For those born in 1930-1939, 15.2% was exposed between the ages of 20 and 29 years. For each age class born in 1900-1939, the proportion exposed at least once by 60 years of age ranged from 18.2% to 24.5% and, of those exposed, the cumulative duration of exposure ranged from 11.3 to 15.4 years by the age of 60 years. A population exposure index showed that the heaviest exposure occurred between 1960 and 1970 and that the age classes born between 1920 and 1929 were the most heavily exposed. Time trends showed that the mean value of this index for the men aged 20-59 years reached a peak in the 1960s and then decreased. Conclusion This study presents data of reasonable validity about occupational asbestos exposure in France and its trends over the past century; the data are being used to forecast the development of male mortality from mesothelioma in France.

Objective Disease activity has been defined using a self-administered instrument, focusing on fatigue, axial pain, peripheral pain, enthesopathy and morning stiffness [Bath Ankylosing Spondylitis Disease Activity Index (BASDAI)]. This validated instrument is simple and takes 40 s to complete, but whether the index is an accurate expression of the component parts, or whether additional weighting would enhance its efficacy, is unclear. Methods Four hundred and seventy-three patients with ankylosing spondylitis received placebo or active non-steroidal anti-inflammatory drug (NSAID) for 6 weeks, and changes between entry and completion were captured by BASDAI and the individual components. Principle component analysis (PCA) was used to explore the best combinations of variables in decreasing order of explained total dispersion and to assess whether a single sum (or algebraic expression) best defined disease activity status. Results At entry, the correlation between BASDAI and the first axis was 0.99, 0.11 with the second, and zero thereafter. Data at study end and relating to change revealed a 100% correlation (R = 1) between the first axis and the sum, with zero for the remainder. Conclusions The data support BASDAI as being a valid and appropriate composite to define disease activity in ankylosing spondylitis. Developed as a simple sum of its components, BASDAI has excellent content validity.

Background Major life changes may play a causative role in health through lifestyle factors, such as alcohol. The objective was to examine the impact of stressful life events on heavy alcohol consumption among French adults. Methods Trajectories of excessive alcohol consumption in 20,625 employees of the French national gas and electricity company for up to 5 years before and 5 years after an event, with annual measurements from 1992. We used repeated measures analysis of time series data indexed to events, employing generalized estimating equations. Results For women, excessive alcohol use increased before important purchase (p = 0.021), children leaving home (p

International audience; Objectives This study aims to estimate the association between repeated measures of occupational risk factors and the incidence of lateral epicondylitis in a large working population.Methods A total of 3710 workers in a French region were included in 2002–2005, and among them 1046 had a complete follow-up in 2007–2010. At both stages, occupational health physicians assessed the presence of lateral epicondylitis and workers self-reported their occupational exposures. Poisson models were performed to assess the incidence rate ratios (IRR) separately by sex using multiple imputed data.Results The annual incidence rate of lateral epicondylitis was estimated as 1.0 [95% confidence interval (95% CI) 0.7–1.3] per 100 workers among men and 0.9 (95% CI 0.6–1.3) among women. Workers aged >45 years had higher incidence than those aged 2 hours/day) was a risk factor, with an age-adjusted IRR of 3.2 (95% CI 1.5–6.4) for workers exposed at both questionnaires [3.3 (95% CI 1.4–7.6) among women].Conclusions This study highlights the importance of temporal dimensions for occupational risk factors on the incidence of lateral epicondylitis. Further research should evaluate the risk associated with the duration and repetition of occupational exposure on the incidence of lateral epicondylitis.

Objectives To show how reweighting can correct for unit nonresponse bias in an occupational health surveillance survey by using data from administrative databases in addition to classic sociodemographic data. Study Design and Setting In 2010, about 10,000 workers covered by a French health insurance fund were randomly selected and were sent a postal questionnaire. Simultaneously, auxiliary data from routine health insurance and occupational databases were collected for all these workers. To model the probability of response to the questionnaire, logistic regressions were performed with these auxiliary data to compute weights for correcting unit nonresponse. Corrected prevalences of questionnaire variables were estimated under several assumptions regarding the missing data process. The impact of reweighting was evaluated by a sensitivity analysis. Results Respondents had more reimbursement claims for medical services than nonrespondents but fewer reimbursements for medical prescriptions or hospitalizations. Salaried workers, workers in service companies, or who had held their job longer than 6 months were more likely to respond. Corrected prevalences after reweighting were slightly different from crude prevalences for some variables but meaningfully different for others. Conclusion Linking health insurance and occupational data effectively corrects for nonresponse bias using reweighting techniques. Sociodemographic variables may be not sufficient to correct for nonresponse.

Aims Previous studies do not provide an accurate estimate of the burden of diabetes on sickness absence. The aim of this study was to measure the impact of diabetes onset on absenteeism starting from the earliest occurrence of the disease. Methods The authors used data from a subsample of the French GAZEL cohort of 506 employees with incident diabetes and 2530 matched diabetes-free participants. Medically certified sickness absence data were obtained from company records (1989–2007). Number of sickness absence days and incidence rates of overall and cause-specific absence spells were compared according to diabetes status across three 5-year periods ranging from 10 years before to 5 years after onset of cases' diabetes. Results The mean number of sickness absence days was persistently higher in participants with diabetes compared with those without diabetes. This difference increased from 16.4 days (95% confidence interval 7.2–25.5) during the 5-year period preceding diabetes onset to 28.5 days (95% CI 16.1–40.9) during the following 5-year period (P = 0.04). This was due to a steeper relative increase in the incidence of long (but not short) absence spells in participants with diabetes versus those without diabetes [incidence rate ratios 1.33 (95% CI 1.08–1.64) and 1.75 (95% CI 1.43–2.14), respectively; P = 0.02]. Diabetes onset was associated with increased rates of circulatory and metabolic absence spells. Conclusions Onset of diabetes is associated with a substantial increase in sickness absence. This suggests that in addition to its burden on work cessation, diabetes weighs heavily on working ability among people who manage to remain employed.

International audience; BACKGROUND: Data on the association between dietary patterns and depression are scarce. The objective of this study was to examine the longitudinal association between dietary patterns and depressive symptoms assessed repeatedly over 10 years in the French occupational GAZEL cohort. METHODS: A total of 9,272 men and 3,132 women, aged 45-60 years in 1998, completed a 35-item Food Frequency Questionnaire (FFQ) at baseline. Dietary patterns were derived by Principal Component Analysis. Depressive symptoms were assessed by the Center for Epidemiologic Studies Depression scale (CES-D) in 1999, 2002, 2005 and 2008. The main outcome measure was the repeated measures of CES-D. Longitudinal analyses were performed with logistic regression based on generalized estimating equations. PRINCIPAL FINDINGS: The highest quartile of low-fat, western, high snack and high fat-sweet diets in men and low-fat and high snack diets in women were associated with higher likelihood of depressive symptoms at the start of the follow-up compared to the lowest quartile (OR between 1.16 and 1.50). Conversely, the highest quartile of traditional diet (characterized by fish and fruit consumption) was associated with a lower likelihood of depressive symptoms in women compared to the lowest quartile, with OR = 0.63 [95%CI, 0.50 to 0.80], as the healthy pattern (characterized by vegetables consumption) with OR = 0.72 [95%CI, 0.63 to 0.83] and OR = 0.75 [95%CI, 0.61 to 0.93] in men and women, respectively. However, there was probably a reverse causality effect for the healthy pattern. CONCLUSION: This longitudinal study shows that several dietary patterns are associated with depressive symptoms and these associations track over time.

Background: Obesity is a contemporary epidemic that does not affect all age groups and sections of society equally. Objective: The objective was to examine socioeconomic differences in trajectories of body mass index (BMI; in kg/m2) and obesity between the ages of 45 and 65 y. Design: A total of 13,297 men and 4532 women from the French GAZEL (Gaz de France Electricité de France) cohort study reported their height in 1990 and their weight annually over the subsequent 18 y. Changes in BMI and obesity between ages 45 and 49 y, 50 and 54 y, 55 and 59 y, and 60 and 65 y as a function of education and occupational position (at age 35 y) were modeled by using linear mixed models and generalized estimating equations. Results: BMI and obesity rates increased between the ages of 45 and 65 y. In men, BMI was higher in unskilled workers than in managers at age 45 y; this difference in BMI increased from 0.82 (95% CI: 0.66, 0.99) at 45 y to 1.06 (95% CI: 0.85, 1.27) at 65 y. Men with a primary school education compared with those with a high school degree at age 45 y had a 0.75 (95% CI: 0.51, 1.00) higher BMI, and this difference increased to 1.32 (95% CI: 1.03,1.62) at age 65 y. Obesity rates were 3.35% and 7.68% at age 45 y and 9.52% and 18.10% at age 65 y in managers and unskilled workers, respectively; the difference in obesity increased by 4.25% (95% CI: 1.87, 6.52). A similar trend was observed in women. Conclusions: Weight continues to increase in the transition between midlife and old age; this increase is greater in lower socioeconomic groups.

International audience; The authors examined the association of 'g' (general intelligence) factor and five specific cognitive measures assessed in 1997-1999 with mortality till 2006 (mean follow-up of 8 years) in the middle-aged Whitehall II cohort study. In age- and sex-adjusted analysis, a decrease in 1 S.D. in memory (hazard ratio (HR), 1.19; 95% confidence interval (CI): 1.02, 1.39) and in Alice Heim 4-I (AH4-I) (HR, 1.16; 95% CI: 1.01, 1.35) was found to be associated with higher mortality. The association with 'g' factor, phonemic and semantic fluency did not reach significance at p

International audience; Average physical stature has increased dramatically during the 20th century in many populations across the world with few exceptions. It remains unclear if social inequalities in height persist despite improvements in living standards in the welfare economies of Western Europe. We examined trends in the association between height and socioeconomic indicators in adults over three decades in France. The data were drawn from the French Decennial Health Surveys: a multistage, stratified, random survey of households, representative of the population, conducted in 1970, 1980, 1991, and 2003. We categorised age into 10-year bands, 25-34, 35-44, 45-54 and 55-64 years. Education and income were the two socioeconomic measures used. The slope index of inequality (SII) was used as a summary index of absolute social inequalities in height. The results show that average height increased over this period; men and women aged 25-34 years were 171.9 and 161.2 cm tall in 1970 and 177.0 and 164.0 cm in 2003, respectively. However, education-related inequalities in height remained unchanged over this period and in men were 4.48 cm (1970), 4.71 cm (1980), 5.58 cm (1991) and 4.69 cm (2003), the corresponding figures in women were 2.41, 2.37, 3.14 and 2.96 cm. Income-related inequalities in height were smaller and much attenuated after adjustment for education. These results suggest that in France, social inequalities in adult height in absolute terms have remained unchanged across the three decades under examination.

Objectives. We examined gender differences in mortality, morbidity, and the association between the 2. Methods. We used health data from 2 studies of middle-aged men and women: the British Whitehall II cohort of employees from 20 civil service departments in London and the 1989 French GAZEL (this acronym refers to the French gas and electric companies) of employees of France's national gas and electricity company. Participants were aged 35 to 55 years when assessed for morbidity and followed up for mortality over 17 years. Results. Male mortality was higher than female mortality in Whitehall II (hazard ratio [HR] = 1.56; 95% confidence interval [CI] = 1.28, 1.91) and the GAZEL cohort (HR = 1.99; CI = 1.66, 2.40). Female excess morbidity was observed for some measures in the Whitehall II data and for 1 measure in the GAZEL data. Only self-reported sickness absence in the Whitehall II data was more strongly associated with mortality among men (P = .01). Conclusions. Mortality was lower among women than among men, but morbidity was not consistently higher. The lack of gender differences in the association between morbidity and mortality suggests that this is not a likely explanation for the gender paradox, which refers to higher morbidity but lower mortality among women than among men.

International audience; BACKGROUND: Common carotid artery intima-media thickness (IMT) is a measure of generalized atherosclerosis and has been shown to be associated with cognitive function. We examine two questions: does socioeconomic status (SES) moderate this association and is IMT more strongly associated with specific aspects of cognitive function? METHODS: Data are drawn from the Phase 7 (2003-2004) of the Whitehall II study (N=3896). In cross-sectional analyses the association between IMT and six measures of cognition (short-term verbal memory, inductive reasoning, vocabulary, semantic and phonemic fluency and a measure of global cognitive status) was examined in analyses adjusted for previous history of coronary heart disease, health behaviours and other vascular risk measures such as blood pressure, cholesterol and body mass index. RESULTS: The overall association between IMT and the six measures of cognition was restricted to the low SES group (p=0.02). Within this group, IMT was significantly associated with inductive reasoning (p=0.001), vocabulary (p=0.002), phonemic (p=0.006) and semantic fluency (p=0.02). The covariates examined explained about a quarter of the association between IMT and cognition in the low SES group. The associations with the measure of inductive reasoning (p=0.02), vocabulary (p=0.02) and phonemic fluency (p=0.04) remained after adjustment for all covariates. CONCLUSIONS: SES is an important modifier of the association between IMT and cognition, an inverse association between the two was observed only in the low SES group. It is possible that high cognitive reserve among the high SES individuals prevents the functional manifestations of atherosclerosis. Verbal memory was not one of the cognitive domains associated with IMT.

International audience; OBJECTIVES: To examine the association between self assessed driving while sleepy and the risk of serious road traffic accidents (RTAs). DESIGN: Prospective cohort study. SETTING: France. PARTICIPANTS: 13 299 of the 19 894 living members of the GAZEL cohort, workers and recent retirees of a French national utility company followed up since 1989. MAIN OUTCOME MEASURES: Frequency of driving while sleepy in the previous 12 months, reported in 2001; rate ratios for serious RTAs in 2001-3, estimated by using generalised linear Poisson regression models with time dependent covariates. RESULTS: The risk of serious RTAs increased proportionally with the frequency of self reported driving while sleepy. After adjustment for sociodemographic characteristics, driving behaviour variables, work conditions, retirement, medical conditions and treatments, depressive symptoms, and sleep disorders, the adjusted rate ratios of serious RTAs for participants who reported driving while sleepy in the previous 12 months "a few times" or "once a month or more often" were 1.5 (95% confidence interval 1.2 to 2.0) and 2.9 (1.3 to 6.3) respectively compared with those who reported not driving while sleepy over the same period. These associations were not explained by any reported sleep disorders. CONCLUSIONS: Self assessed driving while sleepy was a powerful predictor of serious RTAs, suggesting that drivers' awareness of their sleepiness while driving is not sufficient to prevent them from having RTAs. Messages on prevention should therefore focus on convincing sleepy drivers to stop driving and sleep before resuming their journey.

To determine whether systemic markers of bone, cartilage, and synovium can predict structural progression of osteoarthritis (OA).Patients with painful hip OA were treated with diacerein or placebo in a multicentre, prospective, double blind, 3 year follow up trial. The following information was collected at entry: demographics, characteristics of hip OA, and 10 markers: N-propeptides of collagen types I and III, cartilage oligomeric matrix protein, YKL-40, hyaluronan (sHA), matrix metalloproteinases-1 and -3, C reactive protein, C-terminal crosslinking telopeptides of collagen types I and II (uCTX-II). Radiographs were obtained at entry and every year. Structural progression was defined as a joint space decreaseor =0.5 mm or requirement for total hip replacement. Grouped survival analysis was performed with time to structural progression as dependent variable, and clinical data, radiographic findings, treatment groups (diacerein versus placebo), and markers as explanatory measures.In the 333 patients in whom all markers were measured, high functional impairment, a joint space width2 mm, and lateral migration of the femoral head at baseline increased the risk of progression, but diacerein had a protective effect (relative risk = 0.75; 95% confidence interval (CI) 0.54 to 0.96). In addition, patients in whom uCTX-II and sHA were in the upper tertile had a relative risk of progression of 3.73 (95% CI 2.48 to 5.61) compared with patients with markers in the two lower tertiles.In this large cohort, combined measurements of uCTX-II and sHA were a new predictor of the structural progression of hip OA.

To investigate the associations of molecular markers of joint tissue turnover with clinical and radiological variables in patients with hip osteoarthritis (OA).Patients of the ECHODIAH trial cohort (60% female; mean age 63 yrs, disease duration 5 yrs) fulfilling the American College of Rheumatology criteria for hip OA were studied. Pain was assessed using a 100 mm visual analog scale, and the presence of night pain and morning stiffness was observed as the index of joint inflammation. Joint space width (JSW) and subchondral bone sclerosis were assessed on hip radiographs. Ten markers were measured, 8 in serum: N-propeptides of collagen type I (PINP) and type III (PIIINP), cartilage oligomeric matrix protein (COMP), YKL-40, hyaluronan (HA), matrix metalloproteases (MMP1 and MMP3), and ultrasensitive C-reactive protein (CRP); and 2 in urine: C-terminal crosslinking telopeptides of collagen type I (CTX-I) and type II (CTX-II). Analyses of 376 patients with measurements of all the markers included principal component analyses to identify independent clusters of markers; followed by stepwise multivariate regressions to determine associations between markers, clinical variables, and radiographic signs of joint damage.Markers could be segregated into independent clusters: CTX-II, PINP, and CTX-I for cartilage degradation and bone turnover; COMP, PIIINP, and HA as potential markers of synovitis; and CRP and YKL-40, which are likely to indicate systemic inflammation; plus MMP1 and MMP3. After adjustment for age, sex, and body mass index, pain was significantly associated with CTX-II (p = 0.0095) and CRP (p = 0.046) and joint inflammation with COMP (p = 0.013). Radiographic signs of joint damage were associated with CTX-II (p = 0.001 for JSW; p = 0.007 for bone sclerosis).This cross-sectional study of OA molecular markers in a large cohort may provide biological evidence of different pathophysiological processes involved in hip OA. Among the markers measured, CTX-II showed the most consistent association with the symptoms and joint damage of OA.

National audience; Objectifs.– L’objectif était d’évaluer l’association entre des mesures répétées de facteurs professionnels et psychosociaux et l’incidence des épicondylites latérales dans une large population de salariés.Méthodes.– Sur les 3710 salariés qui ont été inclus entre 2002 et 2005 dans un programme de surveillance des TMS dans les Pays de la Loire, 1046 ont eu un suivi complet entre 2007 et 2010. À chaque phase, la présence d’épicondylite a été diagnostiquée par un médecin du travail et les salariés ont rempli un auto-questionnaire sur leurs facteurs de risque professionnels. Des modèles de Poisson univariés et multivariés séparés par sexe ont été utilisés pour évaluer les rapports de taux d’incidences (IRR) sur les données avec imputations multiples.Résultats.– Le taux d’incidence des épicondylites latérale était de 1,0 [0,7 ; 1,3] pour 100 salariés-années chez les hommes et 0,9 [0,7 ; 1,3] chez les femmes. L’incidence chez les travailleurs plus âgés (plus de 45 ans) était plus importante que chez les plus jeunes (moins de 30 ans) mais non significativement. L’exposition aux gestes répétitifs (plus de 4 heures/jours) était un facteur de risque significatif dans les analyses univariées. Un facteur de risque combinant l’effort physique important avec des mouvements aux coudes ou des torsions aux poignets (plus de 2 heures/jours) aux deux questionnaires était fortement associé à l’incidence (IRR ajusté sur l’âge et la répétition « toujours exposé » versus « jamais exposé » de 3,0 [1,5 ; 6,0] chez les hommes et de 3,2 [1,4 ; 7,3] chez les femmes). Le faible soutien social n’était pas associé à l’incidence dans cette étude.Conclusions.– Cette étude souligne l’importance de l’effet des risques professionnels tout au long de la carrière sur l’incidence des épicondylites, en particulier les gestes répétitifs et l’effort physique combiné à des mouvements aux coudes. De prochaines études pourront évaluer l’influence de la durée et de la répétition des expositions professionnelles sur les épicondylites latérales avec des intervalles de mesures pertinents.

BACKGROUND:: Personal responses to stressful life events are suspected of increasing the risk of serious traffic accidents. METHODS:: We analyzed data from a French cohort study (the GAZEL cohort), including a retrospective driving behavior questionnaire, from 13,915 participants (10,542 men age 52-62 years and 3373 women age 47-62 years in 2001). Follow-up data covered 1993-2000. Hazard ratios for serious accidents (n = 713) were computed by Cox's proportional hazard regression with time-dependent covariates. Separate analyses were also performed to consider only at-fault accidents. RESULTS:: Marital separation or divorce was associated with an increased risk of a serious accident (all serious accidents: hazard ratio 2.9, 95% confidence interval = 1.7-5.0; at-fault accidents: 4.4, 2.3-8.3). The impact of separation and divorce did not differ according to alcohol consumption levels. Other life events associated with increased risk of serious accident were a child leaving home (all accidents: 1.2, 0.97-1.6; at-fault accidents: 1.5, 1.1-2.1), an important purchase (all accidents: 1.4, 1.1-1.7; at-fault accidents: 1.6, 1.2-2.1), and hospitalization of the partner (all accidents: 1.4, 1.1-2.0). CONCLUSION:: This study suggests that recent separation and divorce are associated with an increase in serious traffic accidents.

The purpose of this research was to describe women's consumption of alcoholic beverages as it related to their marital status over a 5-year period (1992 through 1996) and to study alcohol consumption around the time of marriage or divorce.The study sample comprised 4,782 women who worked for Electricité de France-Gaz de France and belonged to the GAZEL cohort. The relevant variables collected by five successive annual questionnaires included marital status and alcohol consumption characteristics. Marginal models were used, reflecting the fact that the data were not independent.Divorcees and widows drank less than married women as measured in fewer glasses per day and fewer days per week drinking wine. Women in the oldest generation drank more than the younger women. Getting married was accompanied by an increased level of drinking, especially of wine, beginning a year before the wedding and lasting until 4 years after it. Consumption declined briefly during the year after a divorce.These results may be useful for designing prevention programs aimed at groups of women in the general population in France.

Objective: To evaluate sex differences in the clinical and structural presentation, and natural history of hip OA. Methods: A multicentre, prospective, longitudinal, five year follow up study of 508 patients (302 women, 206 men, mean age 63 (7) years) with painful hip OA. Data collected were baseline demographics, symptomatic, therapeutic, and structural variables; symptomatic variables and changes in joint space width (JSW) during the first year's follow up; requirement for total hip arthroplasty (THA) between the end of the first and fifth years. Statistical analysis: evaluation of sex differences (a) at baseline, in the main characteristics of hip OA using multivariate logistic regression; (b) during the first year of follow up, in the radiological progression of the disease; (c) during the five years of follow up, in the requirement for THA using Kaplan-Meier curves and the log rank test, and of the parameters related to THA, using a multivariate Cox analysis. Results: At entry, women presented more frequently than men with polyarticular OA (mean (SD) articular score 306 (162) v 235 (127)), and superomedial migration of the femoral head (40% v 19%), and had more severe symptomatic disease (patient's overall assessment 46 (23) v 40 (26)). The change in JSW did not differ between women and men after one year, but a greater proportion of women had rapid structural progression (OR=2.34, 95% CI 1.1 to 5.2). THA was performed more often in women. Multivariate analysis suggested that the decision to perform surgery was related more closely to the symptomatic and structural severity of the disease than to the sex of the patient. Conclusion: Hip OA in women is more frequently part of a polyarticular OA, and displays greater symptomatic and structural severity.

STUDY DESIGN Individuals with low back pain were classified by cluster analysis of their responses to the Dallas Pain Questionnaire. These results enabled development of an alternative simple classification tool that yielded results close to those obtained by the cluster analysis. OBJECTIVES To use the Dallas Pain Questionnaire to classify workers with low back pain, and to develop a practical classification tool for physicians. SUMMARY OF BACKGROUND DATA The severity of low back pain is important information for patient management. Questionnaires exploring the multidimensional aspect of low back pain have been used to classify individuals with low back pain. The Dallas Pain Questionnaire is a short self-administered questionnaire that explores the functional and emotional aspects of chronic low back pain. It was tested in a population still at work using a simplified scoring system. METHODS The study included 765 persons with low back pain still at work who completed the self-administered questionnaire. Of these patients, 732 could be classified by cluster analysis according to their answers to the Dallas Pain Questionnaire. A classification tool was developed to categorize them. Using external variables, the groups thus formed were compared. RESULTS Respondents were classified into four classes: class 1 (minor disability), class 2 (intermediate disability), class 3 (disability for some physical activities), and class 4 (physical disability and emotional disturbance). External pain-related variables were clearly associated with increasing low back pain severity as measured by this classification tool. CONCLUSIONS The Dallas Pain Questionnaire discriminated between different groups of persons with low back pain. The proposed classification uses a short, simple practical tool to assess different levels of low back pain.

Objective. To determine a cut-off point above which a change in joint space width (JSW) could be considered as relevant in patients with hip osteoarthritis (OA) on the basis of predicted need for subsequent total hip arthroplasty (THA). Methods. A multicentre, prospective, longitudinal, 5-yr follow-up study was performed. A pelvic radiograph was obtained at entry and after 1 and 2 yr. For each film, the narrowest JSW was measured using a 0.1 mm graduated magnifying glass. The absolute and relative differences between baseline and 1 and 2 yr of follow-up were calculated. We determined the cut-off points above which an absolute or relative decrease in JSW between baseline and 1 and 2 yr of follow-up could be considered relevant on the basis of the predicted need for THA during the remaining years of the study. The need for THA was categorized as ‘yes’ or ‘no’. Thereafter, for each observed change in JSW (0.1 per 0.1 mm or 1% per 1%), the sensitivity and specificity for subsequent THA were calculated. The choice of cut-off was based on maximal sensitivity and specificity, using the graphic representation of correct classification probabilities. In this way it was possible to obtain the best measured JSW threshold with maximal true positive and minimal false positive results. Results. A total of 423 and 385 patients met the criteria for analysis using the decrease in JSW between baseline and 1 and 2 yr respectively. The best cut-off points were absolute decreases in JSW of 0.2 and 0.4 mm and relative decreases in JSW of 15 and 20% after 1 and 2 yr respectively, with corresponding ranges of sensitivity and specificity of 68‐75 and 67‐78%. Conclusion. This work determined the cut-off above which a change in JSW could be considered clinically relevant in patients with hip OA, on the basis of predicted subsequent need for THA. For validation, similar studies should be conducted in other countries with different health-care systems.

Aim. To determine the minimum clinically important difference (MCID) in joint space width (JSW) progression in patients with hip osteoarthritis (OA), based upon evaluation by a panel of clinical experts as a gold standard. Methods. A sample of 298 patients with hip OA was selected from a multicentre, prospective, longitudinal, 3-yr follow-up study. A pelvic radiograph was obtained at entry and after 3 yr. For each film, the narrowest JSW was measured using a 0.1-mm graduated magnifying glass. The difference between baseline and 3-yr follow-up JSW was calculated. Two senior rheumatologists, who were experts in osteoarthritis, evaluated each pair of films and noted whether a clinically relevant deterioration in osteoarthritis stage occurred at 3 yr compared with baseline. Interobserver reliabilities were evaluated using the k coefficient and proportions of agreements. Then, for each measured difference in JSW (0.1 mm per 0.1 mm), the sensitivity and specificity for MCID, defined as the assessment of expert 1, expert 2 or a combination of both, were calculated. This allowed us to obtain, from graphic representations of the correct classification probabilities, the best measured JSW threshold, with the maximal true positive and the minimal false positive results. Results. The mean measured change in JSW was 2 0.63" 0.74 mm. Experts 1 and 2 considered the decrease in JSW to be clinically relevant in 122 (40.9%) and 100 pairs (33.6%) respectively. The proportion of agreements between the experts was 79.9%, with a k coefficient of 0.572. The best measured JSW threshold was 2 0.4 mm for expert 1, expert 2 and the combination of both; sensitivity and specificity were 0.75 and 0.8, 0.71 and 0.72, and 0.75 and 0.7 respectively. Conclusion. This study suggests that a change of at least 0.4 mm in the radiological JSW could be considered clinically relevant. Other studies using other sets of patients and other methods are needed for validation.

OBJECTIVES—Previous projections of mortality from mesothelioma among French men have used the age-generation method, based on the Poisson regression model. In this study an alternative method to model mortality from mesothelioma was used to predict its future trend: this method was based on the risk function that links this mortality to past exposure to asbestos, combined with population exposure data. METHOD—Data on past French asbestos imports were used to model the overall past exposure to asbestos in men and assess two extreme scenarios (optimistic and pessimistic) for its future trends. The number of male deaths occurring between the ages of 50 and 79, from 1997-2050, was then calculated with the risk function for mesothelioma. RESULTS—The results showed that mortality from mesothelioma among French men aged 50-79 will continue to increase, reaching a peak averaging between 1140 (optimistic scenario) and 1300 deaths (pessimistic scenario) annually around the years 2030 and 2040, respectively. No preventive measures applied now will affect this trend before then. These results are similar to those of two other predictions of mortality from mesothelioma among French men: a peak around 2030 of 800-1600 deaths annually among men aged 25-89 years, and a peak around 2020 of 1550 deaths annually among men aged 40-84. CONCLUSIONS—Our results indicate that between 1997 and 2050, the most optimistic and pessimistic trends of future exposure will lead to the deaths from mesothelioma of between 44 480 and 57 020 men, with a corresponding loss of from 877 200 to 1 171 500 person-years of life. Keywords: mesothelioma; risk function; mortality trends; prediction; France

OBJECTIVE: To test whether psychosocial factors at work are predictors of rates of sickness absence. METHODS: The study population consisted of middle aged men and women employed by the French national electricity and gas company (EDF-GDF) in various occupations and followed up since 1989 by annual self administered questionnaires and independent data obtained from the medical and personnel departments of EDF-GDF. The 1995 questionnaire provided information about three psychosocial work factors: psychological demands, decision latitude, and social support at work. Sick-ness absence data were provided by the company's social security department. The occurrence of spells and days of absence in the 12 months after completion of the 1995 questionnaire was studied. Potential confounding variables were age, smoking, alcohol, and marital status, assessed in the 1995 questionnaire, and educational level and occupation, assessed from data provided by the personnel department. This study was restricted to the 12,555 subjects of the initial cohort who were still working and answered the self administered questionnaire in 1995. RESULTS: Low levels of decision latitude were associated with more frequent and longer sickness absences among men and women. Low levels of social support at work increased the numbers of spells and days of absence among men only. These associations weakened after adjustment for potential confounding factors, but remained significant. CONCLUSION: The study indicates that psychosocial factors at work, especially decision latitude, are predictive of sickness absence.

Background: Several scoring systems have been proposed in order to quantify the degree of cartilage damage observed by arthroscopy of the knee in patients with osteoarthritis.Objective: To evaluate the inter-observer reliability of five different scoring systems of arthroscopic evaluation for chondropathy in osteoarthritis of the knee and to evaluate the utility of a training session between different observations on these scoring systems.Methods: Videotapes of knee arthroscopies on five patients with osteoarthritis demonstrating different levels of severity of cartilage damage of the medial tibiofemoral compartment were analyzed by nine observers prior to (pre-training evaluation) and 2 months after a 6 h training session (post-training evaluation) by the following scoring systems: (1) cartilage deterioration by a 100 mm visual analogue scale (VAS), (2) overall assessment of degeneration in the entire medial compartment (cartilage, meniscus, osteophyte) using a 100 mm VAS, (3) French Society of Arthroscopy (SFA) Scoring System, (4) SFA Grading System, (5) American College of Rheumatology (ACR) Scoring System.Results: At the pre-training evaluation, the SFA grading system produced the highest coefficient of reliability (r=0.94), the other systems recording levels of 0.80 for four of the five scoring systems, with lack of improvement in the ACR Scoring System.Conclusion: There was an improved and acceptable inter-observer reliability for at least 2 months follow-up in four of five evaluated scoring systems of arthroscopically graded osteoarthritis of the knee following a training session. A scoring system using a 100 mm VAS may produce the best inter-observer reliability. These results show that scoring chondropathy is possible and demonstrate the importance of training in the analysis of articular cartilage breakdown.

OBJECTIVES: To determine a cut off value for changes in radiological joint space width that allowed definition of radiological progression of hip osteoarthritis not related to measurement method errors and, thereafter, to determine factors predictive of radiological progression of hip osteoarthritis and to evaluate the correlations between clinical and radiological parameters. METHODS: A prospective, longitudinal (one year duration), multicentre study was made of patients with osteoarthritis of the hip (American College of Rheumatology criteria). Data on clinical activity (pain, functional impairment), demographic data (age, gender, body mass index), and femoral head migration (superolateral, superomedial, concentric) were collected when the patient entered the study; radiological grade (joint space width in millimetres at the narrowest point using a 0.1 mm graduated magnifying glass, evaluated by a single observer unaware of the chronology of the films) was recorded at the patient's entry to the study and after one year. RESULTS: Analysis of the means of the differences between two analyses performed by a single observer of 30 pairs of radiographs (one performed after an interval of one year) (0.06 (SD 0.23)) suggested that a change of more than 0.56 mm (2 SD) after a one year follow up could define progression of osteoarthritis of the hip. Of the 508 patients recruited, 461 (91%) completed the one year follow up and radiological progression was observed in 102 (22%). The factors predictive of radiological progression that were identified in the multivariate analysis were: radiological joint space width at entry < or = 2 mm, superolateral migration of the femoral head, female gender, Lequesne's functional index > 10, age at entry > 65 years (odds ratios 2.11, 4.25, 2.51, 2.66, 1.90, respectively). The level of clinical parameters (pain, functional impairment) and the amount of symptomatic treatment required (non-steroidal anti-inflammatory drugs and analgesic intake) accounted for 20% (p < 0.0001) of the variability of the changes in radiological joint space width over the one year study period. CONCLUSION: These data suggest that radiological progression of hip osteoarthritis could be defined by a change in joint space width of at least 0.6 mm after a one year follow up period, is correlated with the changes in clinical status of the patients, and is related not only to demographic data (age, gender), but also to some specific characteristics of osteoarthritis (localisation, radiological severity, clinical activity).