High impact of COVID-19 outbreak in a nursing home in the Nouvelle-Aquitaine region, France, March to April 2020.

BACKGROUND: Elderly people in nursing homes are particularly vulnerable to COVID-19 due to their age, the presence of comorbidities, and community living. On March 14, 2020, at the beginning of the first epidemic wave of COVID-19 in France, a cluster was reported in a nursing home in the Nouvelle-Aquitaine region. We monitored the outbreak as well as the infection prevention and control (IPC) measures implemented. METHODS: A confirmed case was defined as laboratory-confirmed COVID-19 in a resident or staff member present in the nursing home between March 7 and May 1, 2020; and a probable case as a person presenting an acute respiratory illness after contact with a confirmed case. Symptomatic inpatient residents and symptomatic staff members were systematically tested for SARS-CoV-2. In addition, two screening sessions were held on site. RESULTS: We identified 109 cases (98 confirmed, 11 probable). The attack rate was 66% among residents and 45% among staff. Half of all cases were identified during the screening sessions. One-quarter of cases had minor symptoms or were asymptomatic. The case fatality rate among residents was 29%. IPC measures were rapidly implemented such as the quarantine of residents, the reinforcement of staff personal protective equipment, and home quarantine of staff testing positive, which were supplemented in April by systematic controls at the entrance of the nursing home and the creation of additional staff break rooms. CONCLUSIONS: This outbreak confirmed the considerable health impact of SARS-CoV-2 transmission in a nursing home. In addition to the implementation of IPC measures, the early detection of cases through the screening of residents and staff is essential to identify asymptomatic and pre-symptomatic cases and limit the spread of the virus.

Investigation of a leptospirosis outbreak in triathlon participants, Réunion Island, 2013.

We report herein the investigation of a leptospirosis outbreak occurring in triathlon competitors on Réunion Island, Indian Ocean. All participants were contacted by phone or email and answered a questionnaire. Detection and molecular characterization of pathogenic Leptospira was conducted in inpatients and in rodents trapped at the vicinity of the event. Of the 160 athletes competing, 101 (63·1%) agreed to participate in the study. Leptospirosis was biologically confirmed for 9/10 suspected cases either by real-time PCR or serological tests (MAT or ELISA). The total attack rate, children's attack rate, swimmers' attack rate, and the attack rate in adult swimmers were respectively estimated at 8·1% [95% confidence interval (CI) 4·3-14·7], 0%, 12·7% (95% CI 6·8-22·4) and 23·1% (95% CI 12·6-33·8). Leptospirosis cases reported significantly more wounds [risk ratio (RR) 4·5, 95% CI 1·6-13], wore complete neoprene suits less often (RR 4·3, 95% CI 1·3-14·5) and were most frequently unlicensed (RR 6·6, 95% CI 2·9-14·8). The epidemiological investigation supported that some measures such as the use of neoprene suits proved efficient in protecting swimmers against infection. PCR detection in rats revealed high Leptospira infection rates. Partial sequencing of the 16S gene and serology on both human and animal samples strongly suggests that rats were the main contaminators and were likely at the origin of the infection in humans.

Sentinel physician's network in Reunion Island: a tool for infectious diseases surveillance.

UNLABELLED: The surveillance of infectious diseases in Reunion Island is based on a sentinel network of family physicians (FPs) coordinated by the Indian Ocean regional institute for public health surveillance (French acronym OI Cire). The objectives are to identify and monitor outbreaks of influenza, gastroenteritis, and chicken pox, and to characterize circulating influenza viruses. The network can monitor other potentially epidemic diseases. METHOD: The Réunion sentinel network ensures a continuous and permanent surveillance. Physicians send their weekly activity data to the Cire that collects, processes, and interprets it; they also collect samples for biological surveillance of influenza. Statistical thresholds, based on historical data and the estimated numbers of incident cases, are calculated to follow the trend, detect outbreaks, and quantify their impact. RESULTS: The network currently includes 56 FPs and pediatricians, accounting for 6.5% of FPs on the island. The network has clarified the seasonality of influenza during the austral winter and identified the seasonality of acute diarrhea with an epidemic peak when school starts in August. The sentinel FPs's reports allowed monitoring the epidemic trend and estimating the number of cases during the 2005 and 2006 chikungunya outbreaks and 2009 influenza A (H1N1) outbreaks. CONCLUSION: The network has proven its contribution, responsiveness, and reliability for epidemiological surveillance during outbreak. It is an essential tool for infectious diseases surveillance in Reunion Island.

Lessons raised by the major 2010 dengue epidemics in the French West Indies.

Dengue fever has been endemo-epidemic in the whole Region of America. In 2010, Guadeloupe and Martinique experienced historical epidemics, with an estimated attack rate of 10% in two islands. When considering the temporal evolution of epidemiological indicators, an unusual increase in the number of dengue cases could be detected very early. Two main factors might have facilitated the settlement of a viral transmission despite the dry season: a low immunity of the population against the circulating serotype and particular climatic conditions, notably very high temperatures which could have improved both virus and vector efficiency. This unusual situation was considered as a warning sign, and indeed led to major outbreaks in both islands a few weeks later. This event underlines that follow-up of epidemiological indicators is necessary to detect the unusual situations as soon as possible. Furthermore, development of biological and modelling tools should be promoted, as well as integrated management strategies for dengue prevention and control.

Reunion, a sentinel territory for influenza surveillance in Europe.

In Réunion, a French overseas territory located in the southern hemisphere, increase in influenza activity is generally observed several months earlier than in Europe. Influenza activity is monitored in Réunion through a multi-source surveillance system including sentinel practitioners network, hospital emergency department, laboratory and mortality. Since 2009, three successive influenza epidemics occurred on the island. The largest was observed in 2009 while epidemics in 2010 and 2011 were much weaker. In terms of circulating strains, B viruses were predominant at the beginning of the 2009 epidemic but they were completely evicted once A(H1N1)pdm09 circulation started. In 2010, A(H1N1)pdm09 virus was predominant again, but a constant co-circulation of B viruses was observed. In 2011, A(H3N2) virus circulated. The same viruses were identified a few months later in mainland France in the respective seasons. Since 2009, virus circulation, epidemiological trends and health impact of influenza have been similar to those observed in Europe. Influenza surveillance in Réunion may therefore give reliable early information which should be considered apart from the surveillance in mainland France. Then, it might be even a more suitable predictor for Europe than other temperate southern hemisphere countries.

[Epidemiological surveillance du chikungunya on Reunion Island from 2005 to 2011]. / La surveillance epidémiologique du chikungunya a la Réunion entre 2005 et 2011.

INTRODUCTION: The largest chikungunya epidemic in history hit Reunion Island in 2005/6. At that time, a specific monitoring system was set up. Since the end of the epidemic, this monitoring system has continued to operate and has been enhanced. The main objectives of the system are early detection of any case of chikungunya virus infection and epidemiological monitoring. METHODS: During nonepidemic periods, surveillance is focused on active detection of all cases. During epidemic periods, weekly incidence is estimated using data collected via a sentinel physician network. Occurrence of severe and fatal forms is tracked during all epidemiologic situations. RESULTS: In 2005-2006, the estimated number of symptomatic chikungunya-virus infections diagnosed on Reunion Island was 266 000, i.e., an attack rate of 34%. A total of 222 severe cases and 44 mother-to-child transmissions were recorded. Since the end of the outbreak, two clusters have been detected in the western region of the island in 2010 and 2011. The second cluster was non-negligible since it involved 164 cases including 112 that were biologically confirmed. CONCLUSION: Based on results recorded from 2005 to 2006, the chikungunya monitoring system appears to have good sensitivity and reactivity. Recent detection of two disease clusters confirms the system's efficacy.

Re-emergence of dengue in Reunion, France, January to April 2012.

Since January 2012, 20 autochthonous cases of dengue virus (DENV) infection have been identified in Réunion. The first cases were detected on the western coast, but the two co-circulating viruses (DENV-1 and DENV-3) seemed to have spread later to different cities of the island. There is a non-negligible risk of increase in viral transmission over the following weeks, so health professionals and public health authorities in Réunion are preparing to face a potential epidemic.

How to explain the re-emergence of chikungunya infection in Reunion Island in 2010?

In March 2010, a new outbreak of chikungunya infection was detected in the west of Reunion Island. An investigation was launched to describe the incident cases occurrence and to raise hypotheses on factors that could explain the occurrence of this outbreak. All probable or confirmed cases detected by the surveillance system in the western area between March 1st and July 2nd, 2010 were included in the investigation. A standardized questionnaire was performed by phone, including sociodemographic, environmental and behaviour data. A total of 74 cases were described (i.e. response rate of 72%). They were mainly women (sex ratio M/F=0.7), all ages were represented. Most of them (76%) resided in a house; 31% had recently moved, including 22% from metropolitan France. They reported to have been more exposed to mosquitoes and to infected patients than during the major epidemic of 2005-2006. In addition, 41% reported to have reduced their protection against mosquitoes. The results suggest that several concomitant factors contributed to this outbreak: the reintroduction of the chikungunya virus in the island, the population characteristics and environmental factors.

[Epidemic of influenza A(H1N1) 2009 in the French overseas territories of the Americas: epidemiological surveillance set up and main results, April 2009-January 2010]. / L'épidémie de grippe A(H1N1) 2009 dans les territoires français des Amériques: dispositif de surveillance mis en place et principaux résultats, avril 2009-janvier 2010.

Guadeloupe, French Guiana, Martinique, St. Martin and St. Barthelemy were the French territories most exposed to the new influenza A(H1N1)v, and adequate epidemiological surveillance tools were promptly developed in order to detect its emergence. The first stage, "containment phase", consisted in detection and management of individual cases. Then, when an autochthonous A(H1N1)v circulation was confirmed, its evolution has been monitored within the whole population, mainly through data collected from sentinel doctors' networks and virological surveillance. This allowed to detect very early the occurrence of epidemics, and to follow their evolution until they were over. Like all the other Caribbean countries, the five French overseas territories were hit by an outbreak of influenza A(H1N1)v. Although they had globally similar characteristics, each epidemic had its specificity in terms of scale and severity. They started between August and September 2009 in four of the five territories, while the last one, St. Barthelemy, was not affected until the end of the year. Attack rate estimates varied from 28 to 70 per 1000 inhabitants according to the territory, and hospitalisation rate varied from 4.3 to 10.3 per 1000 cases. Severity rate didn't reach 1 per 1000 cases in any of the territories. Compared to metropolitan France, the surveillance system presented several strengths, including the pre-existence of both an active sentinel network and an expert committee on emerging diseases in each territory. On the other hand, specific difficulties appeared, notably linked with logistical aspects of virological surveillance and the co-circulation of dengue virus in Guadeloupe and St. Barthelemy. Despite these difficulties, the different tools allowed early detection of the epidemics and follow-up of their evolution. All of them lead to very concordant results, suggesting that they are completely appropriate to monitor a potential new epidemic wave.

Community outbreak of group B meningococcal disease in southwest France--December 2008 to September 2009.

Between December 2008 and September 2009, 11 cases of invasive meningococcal disease (IMD) group B were reported in a 20 km diameter area in the Département Landes, France. Two of them presented with purpura fulminans and one of them died. The strain responsible for this community outbreak was of the clonal complex ST-269.The incidence rate for IMD group B was 3 per 100,000 inhabitants in Landes from week 40 in 2008 to week 40 in 2009; it was the highest in France during that period. The number of cases observed was significantly higher than expected, especially in young adults (standardised incidence ratio: 23.5, p<0.001). A nightclub located in the 20 km diameter area was a possible place of transmission and a prophylaxis recommended for the staff members helped in decreasing the transmission. However, several cases notified later suggested that the bacteria circulated during several months through healthy carriers in the community. This situation prompted increased surveillance of IMD in Landes and medical practitioners were asked to remain vigilant because of the possible emergence of new cases within the following months.

Performance of a syndromic system for influenza based on the activity of general practitioners, France.

BACKGROUND: In France, as in other industrialized countries, syndromic surveillance systems for the early detection of illnesses have proliferated, but few validation studies on these systems performances exist. In Bordeaux, a south-western city in France, a system using a network of general practitioners house calls, such as SOS Médecins, provided local health data used to guide health service response, in particular in case of flu-like pandemic. We explored the capacity of SOS Médecins system to identify and follow influenza outbreaks using data from the Sentinel network, considered as being a gold standard for tracking seasonal influenza in France. METHODS: Data from SOS Médecins were analysed and compared with data from the Sentinel network. The sensitivity and specificity of SOS Médecins system were evaluated for different simulated thresholds. RESULTS: A relationship between the number of visits for influenza from SOS Médecins and the number of influenza cases from the Sentinel network was observed; data from the two systems were highly correlated. We showed the capacity of SOS Médecins system to identify outbreaks with a sensitivity and specificity of 93%. CONCLUSION: The sensitivity and specificity of SOS Médecins for early outbreak detection showed the value of these data in monitoring influenza activity.

Validation of a syndromic surveillance system using a general practitioner house calls network, Bordeaux, France.

A new syndromic surveillance system has been developed in Bordeaux City, South West France, using a general practitioners' house calls network. Routinely collected, sociodemographic data, patients' complaints and medical diagnoses made at the end of the visit were monitored using syndrome groups such as influenza syndromes, bronchiolitis, gastrointestinal, respiratory syndromes and others, based on International Classification of Primary Care (ICPC)-2 codes. A process control chart was implemented in order to distinguish signals of interest from "background noise". In 2005 and 2006, a total of 303,936 visits were recorded. Seasonal epidemics of influenza-like illness, bronchiolitis or gastrointestinal were identified. The automated and real time nature of the system also allowed the early detection of unusual events such as an acute increase in the number of heat syndromes during the heat-wave that occurred in France in July 2006. This new system complements existing surveillance programs by assessing a large part of episodes of illness that do not require hospital admissions or the identification of an etiologic agent. Attributes and advantages of the system, such as timeliness and diagnostic specificity, demonstrated its utility and validity in term of syndromic surveillance purposes, and its extension at the national level is in process.

[Is there an excess of AIDS-related mortality in Aquitaine? Data collected through the French mandatory reporting system, 1982-2005]. / Existe-t-il une surmortalité par sida en Aquitaine? Etude sur les données de déclaration obligatoire, 1982-2005.

OBJECTIVES: The mortality rate among AIDS patients reported through the French mandatory reporting system since 1982 is particularly high in Aquitaine, a south west area of France. This study was made to confirm that mortality was higher in Aquitaine patients than in the rest of France, and to determine its causes. DESIGN: All the cases reported between January 1982 and June 2005 were included. Cox's proportional hazard models were used to determine the relative risk (RR) of death among patients living in Aquitaine. Two periods were considered to take into account the introduction of highly active antiretroviral therapy (HAART) in 1996. RESULTS: Since 1982, 60,212 AIDS patients were reported. After adjustment on the main patient features, the risk of death was significantly higher in AIDS patients living in Aquitaine than in other regions (RR=1.12; IC95%=[1.06-1.18]) before the introduction of HAART. After 1996, this difference did not persist, and an opposite trend was even observed (RR=0.90; IC95%=[0.75-1.09]). CONCLUSIONS: The higher overall mortality rate in Aquitaine was due to a higher risk of death in that area before 1996. Several hypotheses can explain this result, such as a less exhaustive notification, or different characteristics of the subjects. The opposite trend that was observed after the introduction of HAART will have to be checked in the next few years.

Short-term associations between fine and coarse particles and hospital admissions for cardiorespiratory diseases in six French cities.

OBJECTIVES: Little is known about the potential health effects of the coarse fraction of ambient particles. The aim of this study is to estimate the links between fine (PM(2.5)) and coarse particle (PM(2.5-10)) levels and cardiorespiratory hospitalisations in six French cities during 2000-2003. METHODS: Data on the daily numbers of hospitalisations for respiratory, cardiovascular, cardiac and ischaemic heart diseases were collected. Associations between exposure indicators and hospitalisations were estimated in each city using a Poisson regression model, controlling for confounding factors (seasons, days of the week, holidays, influenza epidemics, pollen counts, temperature) and temporal trends. City-specific findings were combined to obtain excess relative risks (ERRs) associated with a 10 mug/m(3) increase in PM(2.5) and PM(2.5-10 )levels. RESULTS: We found positive associations between indicators of particulate pollution and hospitalisations for respiratory infection, with an ERR of 4.4% (95% CI 0.9 to 8.0) for PM(2.5-10 )and 2.5% (95% CI 0.1 to 4.8) for PM(2.5). Concerning respiratory diseases, no association was observed with PM(2.5), whereas positive trends were found with PM(2.5-10), with a significant association for the 0-14-year-old age group (ERR 6.2%, 95% CI 0.4 to 12.3). Concerning cardiovascular diseases, positive associations were observed between PM(2.5) levels and each indicator, although some did not reach significance; trends with PM(2.5-10 )were weaker and non-significant except for ischaemic heart disease in the elderly (ERR 6.4%, 95% CI 1.6 to 11.4). CONCLUSIONS: In accordance with other studies, our results indicate that the coarse fraction may have a stronger effect than the fine fraction on some morbidity endpoints, especially respiratory diseases.

Estimated risk of endocarditis in adults with predisposing cardiac conditions undergoing dental procedures with or without antibiotic prophylaxis.

BACKGROUND: Although antibiotic prophylaxis for infective endocarditis (IE) has been recommended for persons with predisposing cardiac conditions (PCCs) for many years, its efficacy, which has not been demonstrated, has been recently challenged. To assess its usefulness, we estimated the risk of developing IE after undergoing a dental procedure for which subjects would be eligible for prophylaxis, both in subjects having (protected procedure) or not having (unprotected procedure) received antibiotic prophylaxis. METHODS: The number of French persons with PCCs, the annual number of dental procedures in which subjects would be eligible for antibiotic prophylaxis, and the number of procedures that were unprotected were estimated on the basis of a survey performed on a sample of 2805 subjects aged 25-84 years. The annual number of IE cases possibly due to an unprotected procedure was estimated on the basis of a 1-year epidemiological study of IE conducted in an area inhabited by 16 million people. RESULTS: After standardization, extrapolation of results to the age-equivalent general population (39 millions subjects) indicated the following: first, 3.3% (95% confidence interval [CI], 2.6%-4%) of the subjects had PCC, 2.7 million (95% CI, 2.3-3.2 million) of whom had undergone at least 1 at-risk dental procedures within the survey year, and the procedures were unprotected in 62% of cases; second, 37 (95% CI, 18-68; 2.7%) of the 1370 annual IE cases in France were possibly related to unprotected procedures. Thus, the risks of developing IE were estimated to be 1 in 46,000 for unprotected procedures (1 in 10,700 and 1 in 54,300 for subjects with prosthetic and native valve PCC, respectively) and 1 in 150,000 for protected procedures. CONCLUSIONS: A huge number of prophylaxis doses would be necessary to prevent a very low number of IE cases.

Correlates of regular fish consumption in French elderly community dwellers: data from the Three-City study.

OBJECTIVE: Several studies have found an inverse association between fish consumption and risk of dementia. The objective of this study was to describe the characteristics of older fish consumers, in order to identify potential confounders in this relationship. DESIGN: Cross-sectional analysis of the baseline data of a prospective cohort study. SETTING: Bordeaux, Dijon, and Montpellier (France) in 1999-2000. SUBJECTS: A total of 9280 community dwellers aged 65 y and above participating in the baseline examination of the Three-City Study. INTERVENTIONS: All participants answered a face-to-face interview and underwent a physical examination. Cross-sectional analyses were performed by logistic regressions. RESULTS: Regular fish consumers (at least weekly) had a higher education (OR from 1.19 to 1.65, P = 0.0003) and income (OR from 1.37 to 1.89, P < 0.0001). Controlling for age, sex, education, and city, they had a higher consumption of pulses, fruit, and vegetables (P < 0.001). They were more often alcohol drinkers (P < 0.0001). They felt in better health (P < 0.05), exhibited less depressive symptoms (P < 0.001), and scored higher on the Mini Mental Status Examination (P < 0.05). However, their objective physical health status was not better, except that they were leaner. They suffered more often from hypertension and past stroke. CONCLUSIONS: Socioeconomic status, dietary habits, depression, and vascular risk factors could act as confounders in the relationship between fish consumption and risk of dementia.

Sociodemographic differences in dietary habits in a population-based sample of elderly subjects: the 3C study.

PURPOSE: Diet plays an important role in maintaining quality of life in ageing and can be influenced by sociodemographic factors. We aimed to describe dietary habits in a large population-based sample of elderly people and to explore how they may vary according to sex, age, educational level and lifestyle. METHODS: 9250 community dwellers aged 65 years and over recruited in three French cities were interviewed about their food habits with a brief frequency questionnaire. For each food considered, the subjects were classified as occasional or regular consumers. Quantities of alcoholic beverages consumed daily were also assessed. Dietary habits were compared for men and women separately according each sociodemographic factor. RESULTS: Women were characterized by a more regular consumption of fruit and vegetables and drunk fewer alcoholic beverages. Older subjects ate meat, fish, cereals, raw vegetables and pulses less regularly. The proportion of regular consumers of fish, raw fruit, raw vegetables and cooked fruit or vegetables and the quantity of alcohol consumed increased with educational level. Subjects living alone were less regular consumers of almost all foods. DISCUSSION: The results show that even in a sub-population of elderly people, dietary patterns can vary greatly according to several factors. Identifying such factors is of value for future analytic studies on nutrition and health in the elderly. In view of dietary recommendations advising a high consumption of fruit and vegetables, subjects living alone and people with a low educational level are particularly at risk and should be encouraged to eat more balanced meals.

Relationship between body mass index and different domains of disability in older persons: the 3C study.

OBJECTIVE: To study the relationships between body mass index (BMI) and different domains of disability in elderly subjects from the French 3C study. SETTING: Three cities in France: Bordeaux (South-West), Dijon (North-East) and Montpellier (South-East). DESIGN: Cross-sectional study. SUBJECTS: A sample of 8966 elderly community dwellers (age: 65-101 y). MAIN OUTCOME MEASURES: BMI, continence, basic and Instrumental Activities of Daily Living (ADL and IADL) and mobility. Adjustment variables: age, educational level, lifestyle, cognitive functioning, smoking and drinking history, depression, dyspnea, diabetes and indicator of cardiovascular disease. RESULTS: Obesity (BMI> or =30 kg/m2) was significantly associated with disability in each domain for women. The relationship tended to be linear for ADL and for continence; whereas for IADL, underweight women (BMI <21 kg/m2) were also at higher risk of disability. In men, relationships were weaker since BMI was only associated with mobility restriction, with a higher risk for both underweight and obese subjects. CONCLUSION: These results are in favor of a strong association between obesity and the three domains of disability and incontinence. Weaker relationships between underweight and disability were observed. Results suggest that maintaining a BMI in the healthy range could contribute to independence in activities of daily living.

Nutritional factors and risk of incident dementia in the PAQUID longitudinal cohort.

PURPOSE: The aim of this study was to present a summary of the results obtained in the PAQUID study regarding wine consumption, fish and seafood consumption and vitamin A and E plasma concentrations. METHODS: This study was carried out from the PAQUID (Personnes Agees QUID) cohort, an epidemiological study on cognitive and functional aging after 65 years in Southwestern France. The relationship between nutritional factors and risk of incident dementia was measured using logistic regression or Cox proportional hazards models on different subsamples of the cohort. RESULTS: Among 2950 initially non-demented subjects, moderate drinkers had a decreased relative risk of developing a dementia in the subsequent 8 years (RR=0.56) compared to non-drinkers. In a sub-sample of 1416 subjects, those who ate fish or seafood at least once a week had a significant reduced risk of incident dementia adjusted for age and sex (RR=0.66). Lastly, a nested casecontrol study was performed among 182 subjects. Adjusted for confounders, the risk of dementia was significantly increased for the lowest vitamin E concentration compared to the highest one (OR=2.54). DISCUSSION: These results are in agreement with a possible protective role of a diet rich in poly-unsaturated fats and antioxidant components. However, confounding factors such as education and lifestyle may be involved and should be further explored.

Association between antioxidant nutritional indicators and the incidence of dementia: results from the PAQUID prospective cohort study.

OBJECTIVE: To analyse the relation between antioxidant vitamins A, E, and malondialdehyde (MDA) lipoperoxidation product plasma concentrations with incident dementia. DESIGN: : A nested case-control within the PAQUID (Personnes Agées QUID) cohort. SETTING: The PAQUID population-based prospective cohort in southwestern France. SUBJECTS: Among 626 subjects with blood collection at baseline, 46 developed a dementia during the follow-up and were considered to be cases. Each case was matched (on age and sex) to three controls. RESULTS: Plasma vitamin E concentrations were lower among cases (mean value at 22.62 micromol/l (s.d.: 7.38) vs 24.99 (s.d.: 6.73 among controls). The same trend was observed for vitamin A concentrations, but the difference was not significant. On the contrary, MDA concentrations tended to be higher (mean value 1.35 micromol/l (s.d.: 0.53) vs 1.23 (s.d.: 0.44)) among cases. In logistic regression models, plasma values were split into tertiles. Adjusted for confounders, the risk of dementia was significantly increased in the lowest vitamin E tertile (< or =21.0 micromol/l) (OR=3.12, P=0.033) compared to the highest one (> or =25.5 micromol/l). The risk of Alzheimer's disease was also increased, with borderline significance (OR=3.06, P=0.053). Risks associated with vitamin A were nonsignificant. Similarly, there was a trend to an increased risk of dementia in the highest tertile of MDA (OR=1.67, P=0.31). CONCLUSIONS: These results suggest that subjects with low plasma vitamin E concentrations are at a higher risk of developing a dementia in subsequent years.