Greener living environment healthier people?: Exploring green space,physical activity and health in the Doetinchem Cohort Study

BackgroundEvidence is emerging that more green space in the living environment is associated with better health, partly via the pathway of physical activity.ObjectivesWe explored the cross-sectional and longitudinal associations between green space and physical activity and several health indicators in the Doetinchem Cohort Study.MethodsA random sample of men and women aged 20–59 years at baseline was measured max 5 times with 5 year-intervals in the period 1987–2012. Data of round 3–5 were used. Measurements were based on examinations (height, weight, blood pressure) or questionnaires (physical activity, perceived health, depressive symptoms, chronic conditions). The percentage of green space (mainly urban and agricultural green) around the home address (125 m and 1 km) was calculated using satellite data.ResultsMore agricultural green was associated with less time spent on bicycling (β1 km = − 0.15, 95%CL − 0.13; − 0.04) and sports (β1 km = − 0.04, 95%CL − 0.07; − 0.01) and more time spent on gardening (β1 km = 0.16, 95%CL 0.12; 0.19) and odd jobs (β1 km = 0.10, 95%CL 0.05; 0.15), and this was in the other direction for urban green. For only a few of the many health indicators a positive association with green was found, and mainly for total green within 1 km radius. Longstanding green or a transition to more green did not show more pronounced associations with health.ConclusionsFor the green space range of the Doetinchem area the findings do not strongly support the hypothesis that the percentage of green in the living environment affects health positively. The distinction by type of green may, however, be relevant for physical activity.     

Relationship between psychosocial factors at work and incidence of perceived health problems in the GERICOTS cohort

BackgroundThe objective of this study was to investigate the relationship between the incidence of health problems and the psychosocial work environment in a French occupational cohort.MethodsAmong the 2062 employees in the North of France who participated in the GERICOTS survey between 1999 and 2004, 1154 subjects who kept the same full-time shift in the same firm during the study period (797 men and 357 women) were included. Job strain was assessed using Karasek's model–the strain profile (high psychological demand and low decision latitude) and the iso-strain profile (strain profile and low social support)–and Siegrist's model, Effort–Reward Imbalance (ERI) (high effort and low rewards). Perceived health status was assessed using the four dimensions of the Nottingham Health Profile (emotional reactions, sleep, social isolation, pain). The results are given by the odds ratio adjusted for age, occupational category, and size of firm.ResultsAt baseline, higher prevalence of health problems was found in workers with job strain, e.g., between emotional reactions and iso-strain for men, OR = 3.50 (2.19–5.60) and for women, 2.64 (1.39–5.04) or between sleep disorders and ERI for men, OR = 2.41 (1.71–3.40) and for women, OR = 2.41 (1.39–4.20). Longitudinal analysis showed a significant relationship between incidence of health problems and level of job strain in 1999, e.g., between sleep disorder incidence and strain profile, OR = 1.89 (1.16–3.06) and ERI, OR = 2.20 (1.43–3.38).ConclusionThese results show a significant relationship between perceived health and job stress in 1999 but also between incidence of health problems between 1999 and 2004 and job strain in 1999.     

Associations between multiple green space measures and birth weight across two US cities

IntroductionSeveral measures of green space exposure have been used in epidemiological research, but their relevance to health, and representation of exposure pathways, remains unclear. Here we examine the relationships between multiple urban green space metrics and associations with term birth weight across two diverse US cities.MethodsWe used Vital Statistics data to create a birth cohort from 2005 to 2009 in the cities of Portland, Oregon (n = 90,265) and Austin, Texas (n = 88,807). These cities have similar green space levels but very different population and contextual characteristics. Green space metrics derived from mother's full residential address using multiple buffer distances (50–1000 m) included: Landsat Normalized Difference Vegetation Index (NDVI), % tree cover, % green space, % street tree buffering, and access to parks (using US EPA EnviroAtlas Data). Correlation between green space metrics were assessed and mixed models were used to determine associations with term birth weight, controlling for a comprehensive set of individual and neighborhood factors. City-specific models were run to determine how contextual and population differences affected green space associations with birth weight.ResultsWe observed moderate to high degrees of correlation between different green space metrics (except park access), with similar patterns between cities. Unadjusted associations demonstrated consistent protective effects of NDVI, % green space, % tree cover, and % street tree buffering for most buffer sizes on birth weight; however, in fully adjusted models most metrics were no longer statistically significant and no clear patterns remained. For example, in Austin the difference in birth weight for the highest versus lowest quartile of % green space within 50 m was 38.3 g (95% CI: 30.4, 46.1) in unadjusted and −1.5 g (98% CI: −8.8, 6.3) in adjusted models compared to 55.7 g (95%CI: 47.9, −63.6) and 12.9 g (95% CI: 4.4, 21.4) in Portland. Maternal race, ethnicity and education had the largest impact on reducing green space and birth weight associations. However, consistent positive associations were observed for the high density areas of both cities using several green space metrics at small buffer distances.ConclusionsThis study highlights the importance of understanding the individual and contextual factors that may confound and/or modify green space and birth weight associations.     

Social comparison of distress and mental health help-seeking in the US general population

The role of social comparison of distress in the mental health help-seeking process remains largely unexplored. The aim of this study was to examine the association of socially compared distress with mental health help-seeking and perceived need for mental health care in a population sample. In 36,679 adult participants of the 2003 US National Survey on Drug Use and Health, data on 12-month help-seeking and perceived unmet need for care were compared between participants who described themselves as more worried, nervous or anxious than others vs. participants who described themselves as no more worried, nervous or anxious than others. Compared to participants who described themselves as no more worried, nervous or anxious, than others, those who described themselves as more worried, nervous or anxious were significantly more likely to seek professional help (adjusted odds ratio = 1.84) or to perceive an unmet need for such help (adjusted odds ratio = 1.44). It is concluded that social comparison of distress is a significant correlate of mental health help-seeking and perceived unmet need for such help. Individual variations in social comparison of distress may partly explain the discrepancy between need—as measured by non-compared distress—and help-seeking in the general population.     

Impacto sobre el cuidador principal de una intervención realizada a personas mayores dependientes

ObjectiveTo evaluate the impact of joining a home care program on primary caregivers of dependent elderly people.DesignNon-randomised “before-after” intervention study.SettingPrimary Care.ParticipantsPrimary carers of elderly dependent people included in a home care program (n = 156; 7.8% loss to follow up).InterventionsInclusion in a home care program for chronically dependent elderly and the assessment of the primary carer in the same year.Variables assessedperceived health, frequency of visits, questionnaires of quality of life (Nottingham questionnaire), psychological health (Goldberg questionnaire), social support (Duke-UNC scale) and overburden of caregivers (Zarit questionnaire) and satisfaction with care received.ResultsThere were no significant changes in perceived health. Improvement in the areas of energy, sleep, emotional and social relationship of the quality of life. Decreased attendance (8.4 vs. 7.5, p < 0.05) and the percentage of overusers (30.1 vs 6.9%, P < .01). A reduced percentage of caregivers expressed low social support (8.3 vs 2.8%, P < .05) and caregiver overburden (56.4 vs 44.4%, P < .05). 90.3% of caregivers believed that care had improved at the end of intervention, with a significant improvement of satisfaction of overall medical and nursing care received (7.6 vs 8, 4, 7.9 vs 8.5 and 7 vs 8.5; P < .05).ConclusionsJoining a home care program for dependents has a positive impact on their primary caregiver and improves their perception of care received, reducing their use of health services, reducing the level of overburden and their perceived lack of social support.     

Factors affecting uptake of influenza vaccination among family physicians

AimThe aim of this study was to determine the factors that influenced the decisions of family physicians working in primary care health services to receive influenza vaccines.MethodsThis cross-sectional study was performed between June 2014 and September 2014. Physicians were reached electronically via e-mail. A self-reported questionnaire consisting of 50 items covering potential factors that may have influenced their decision to receive vaccination, including perceived risk, severity of the perceived risk, perceived benefit, perceived barriers, cues to action, attitudes, social influences and personal efficacy, was administered to the study participants. Cronbach's alpha for the questionnaire was determined to be 0.92 in the pilot study.ResultsThe response rate was 27.5% (n = 596). Regularly vaccinated physicians accounted for 27.3% of the responses. The median age was 41.84 ± 7.80, and the median working duration of the group was 17.0 ± 7.8 years. The factors that led to increased vaccination compliance (p < 0.05) included working duration, age, chronic disease history and living with a person over 65 years. Nearly all major domains, i.e., perceived risk, severity of the perceived risk, perceived benefit, perceived barriers, attitudes, social influences and personal efficacy, there were differences between the compliant and noncompliant groups. Multi-regression analyses revealed that risk perception, organizational factors such as time and convenient vaccination increased vaccine compliance. However, the perceived necessity to be vaccinated annually had a negative effect on vaccination behaviour (p < 0.05).ConclusionStrategies aimed to increase the flu vaccination ratio among physicians that do not take different factors into account are more likely to be unsuccessful. In the planning and implementation of strategies aiming to increase the vaccination ratio among physicians, it is both necessary and important to take into account behavioural and organizational factors.     

Lonely hearts don't get checked: On the role of social support in screening for cardiovascular risk

ObjectiveRegular cardiovascular risk screening can prevent cardiovascular disease through timely implementation of lifestyle changes or medication. However, few studies have investigated what factors promote regular screening for risk factors like hypertension and high blood cholesterol. The aim of this study was to investigate the relationship between social support and adherence to cardiovascular risk screening.MethodsWe analyzed data from the Spanish National Health Survey—a cross-sectional representative survey conducted by the Spanish Ministry of Health in 2012 (N = 21,007). Participants reported whether they had their blood pressure and cholesterol levels measured by a health professional in the previous 12 months. Social support (i.e., the perception that emotional and practical support was available when needed) was measured with a validated scale. Multiple logistic regressions were conducted adjusted for demographic and health-related factors.ResultsCompared to individuals who reported a lack of social support, individuals who perceived sufficient social support were on average twice more likely to report participation in blood pressure screening, OR = 2.06, 95% CI [1.60, 2.66] and cholesterol screening, OR = 2.85, 95% CI [1.99, 4.09]. These effects were uniform across different demographics and were replicated in a previous wave of the survey. Factors associated with worse screening adherence were low social class, being single or widowed, smoking, alcohol consumption, and no history of cardiovascular risk.DiscussionPerceptions of social support are positively related to cardiovascular risk screening adherence. Future research should investigate what type of social support most effectively increases screening participation among high risk populations.     

Perceptions of social norms and exposure to pro-marijuana messages are associated with adolescent marijuana use

Despite consistent declines in rates of cigarette use among adolescents in the last five years, rates of marijuana use have remained constant, with marijuana being the most widely used illegal drug among adolescents. More work is needed to understand how social norms, perceived risks and benefits, and social media messaging impact use of marijuana. This study compared perceptions and social norms related to marijuana, blunts and cigarettes. Additionally, we assessed how perceptions related to social norms, risks and benefits, and exposure to pro- versus anti-marijuana messaging is related to use. Participants were 786 adolescents from Southern and Northern California (36.7% male, 63.21% females; mean age = 16.1 years; SD = 1.6). Participants came from diverse ethnic backgrounds, with 207 (26.61%) White, 171 (21.98%) Asian/Pacific Islander, 232 (29.82%) Hispanic, and 168 (21.59%) other. Results indicated that marijuana and blunts were consistently perceived as more socially acceptable and less risky than cigarettes (p < 0.01). Participants who reported that their friends used marijuana had a 27% greater odds of using marijuana themselves. Further, seeing messages about the good things or benefits of marijuana use was associated with a 6% greater odds of use [OR 1.06 (CI 1.00, 1.12)]. This study's findings offer a number of important public health implications, particularly as states move towards legalization of marijuana for recreational use. As this occurs, states need to take adolescents' perceptions of risks, benefits, social norms, and peer influences into account as they implement strategies to reduce youth use of marijuana and blunts.     

Estado de salud de personas inmigrantes en Cataluña desde una perspectiva de género: Proyecto PELFI

ObjectiveThe objective of the study is to assess the health status of immigrant men and women from non-EU countries living in the Metropolitan Area of Barcelona (Catalonia, Spain) and to identify the social determinants of health from a gender perspective.MethodCross-sectional analysis from a cohort of immigrant families recruited in Badalona and Santa Coloma de Gramenet, in Spain (PELFI cohort). In 2015–2016, 167 immigrants answered the baseline epidemiological questionnaire and the 5-level EuroQol instrument (EQ-5D) which measures health status. To identify health determinants, Tobit models were constructed to the EQ-5D index.ResultsWomen rated poorer self-perceived health (p = 0.005). To be diagnosed with and illness was only associated with poor self-perceived health among men (p < 0.05). Length of residence, domestic work, and especially double workload deteriorated women's health. After adjusting models by demographics and social determinants, permanent administrative status (?0.136; p = 0.015) and social support (0.182; p = 0.02) were associated with health status in both sexes. To have a job was associated with better health only in men.ConclusionsNon-EU immigrants living in the Metropolitan Area of Barcelona are a socially vulnerable group of population and present inequalities in health by sex. Social support and occupation are key factors of their health status. Interventions to reduce immigrant vulnerabilities and inequalities in health should promote their social inclusion and cohesion from a gender perspective.     

The likelihood of participation in clinical trials can be measured: the Clinical Research Involvement Scales

ObjectiveWe developed the Clinical Research Involvement Scales (CRIS) to assess the willingness to participate in a clinical trial.Study Design and SettingDiverse populations (N = 919) aged 18 years or older from Atlanta, Georgia, were included in comprehensive testing of the 41-item CRIS instrument. The formative phase focused on item content for the new measures (n = 54). Questionnaires from potential vaccine trial participants (n = 865), collected at multiple time points, resulted in the evaluation of scale reliability and validity (i.e., attitudes, behavioral and normative beliefs, perceived social support for clinical research participation, social norm compliance, perceptions of the clinical research organization, and perceived relevance of the research endeavor).ResultsQualitative testing revealed adequate comprehension and content validity of the initial item set. The subjective norms domain (n = 3) initially exhibited poor internal consistency in pilot testing (Cronbach's α = 0.525); yet, rewording of the items resulted in consistently stable measurement improvement (Cronbach's α = 0.850). Each of the CRIS subscales demonstrated extremely high reliability, ranging from 0.734 to 0.918. Confirmatory factor analysis verified item–factor relationships and determined construct and convergent validity (root mean square error of approximation = 0.068; comparative fit index = 0.835).ConclusionsCRIS is a reliable instrument for measuring community attitudes toward participation in biomedical research studies. Results of this study support the use of these scales to recruit diverse populations to clinical trials.     

Élaboration et qualités psychométriques d’une échelle de bien-être au travail. Étude SERENAT auprès de salariés vus en médecine du travail

BackgroundWell-being at work is nowadays a major public health challenge. It includes, among others, absence of psychological (anxio-depressive) symptoms, perceived positive work conditions (environment and organization), happiness and good quality of life at work. Many studies have shown that social support and control at work protect mental health while high job demands and effort-reward imbalance are risk factors for anxiety and depression. There is currently no global indicator to measure both the state of mental health and social working conditions. The main objective of this work is to construct and explore the psychometric properties of scale of well-being at work called “Serenat” in order to validate it.MethodsThe Serenat Scale is a self-report questionnaire composed of 20 items. All items are scored on a four-point Likert scale ranging from 0 (strongly disagree) to 3 (strongly agree) resulting in a range of 0 to 60. It was constructed from data collected from the literature and from consultations in an Occupational Health Unit. From January 2014 to May 2017 193 subjects who have consulted an occupational doctor are included in this cross sectional survey. Validation included item quality and data structure diagnosis, internal consistency, intraobserver reliability evaluation and external consistency.ResultsThe Serenat scale showed very good item quality, with a maximal non-response rate of 0.01 % per item, and no floor effect. Factor analysis concluded that the scale can be considered unidimensional. Cronbach's alpha of internal consistency was 0.89. The intraclass correlation coefficient for intraobserver reliability was 0.89. Serenat scale was correlated with HADS (r = −0.54; P < 0.001), STAI-Y (r = −0.78; P < 0.001) and BDI-13 (r = −0.57; P < 0.001).ConclusionSerenat's well-being at work scale shows good psychometric properties for final validation. It could be useful to occupational physicians for individual and collective screening.Trial RegistrationClinicalTrials.gov ID: NCT02905071.     

Family history and perceived risk of diabetes,cardiovascular disease,cancer, and depression

BackgroundFamily history is a useful and inexpensive tool to assess risks of multifactorial diseases. Family history enables individualized disease prevention, but its effects on perceived risks of various diseases need to be understood in more detail. We examined how family history relates to perceived risk of diabetes mellitus, cardiovascular disease (CVD), cancer, and depression, and whether these associations are independent of or moderated by sociodemographic factors, health behavior/weight status (smoking, alcohol consumption, physical activity, BMI [kg/m²]), or depressive symptoms.MethodsParticipants were Finnish 25–74-year-olds (N = 6258) from a population-based FINRISK 2007 study. Perceived absolute lifetime risks (Brewer et al., 2004; Becker, 1974; Weinstein and Nicolich, 1993; Guttmacher et al., 2004; Yoon et al., 2002) and first-degree family history of CVD, diabetes, cancer and depression, and health behaviors were self-reported. Weight and height were measured in a health examination.ResultsFamily history was most prevalent for cancer (36.7%), least for depression (19.6%). Perceived risk mean was highest for CVD (2.8), lowest for depression (2.0). Association between family history and perceived risk was strongest for diabetes (β = 0.34, P < 0.001), weakest for depression (β = 0.19, P < 0.001). Adjusting for sociodemographics, health behavior, and depressive symptoms did not change these associations. The association between family history and perceived risk tended to be stronger among younger than among older adults, but similar regardless of health behaviors or depressive symptoms.DiscussionAssociation between family history and perceived risk varies across diseases. People's current understandings on heritability need to be acknowledged in risk communication practices. Future research should seek to identify effective strategies to combine familial and genetic risk communication in disease prevention.     

Combination and adaptation of two tools to assess parental feeding practices in pre-school children

ObjectiveTo test a combined version of the Child Feeding Questionnaire and the expanded concept of parental control (“covert” and “overt”) among Portuguese preschool children.MethodsThe final questionnaire comprised 38 items and 9 subscales. The translated questionnaire was self-administered to 854 mothers of 4 year-old children from the Generation XXI birth cohort. Maternal and children's weight and height and socio-demographic characteristics were measured. The global goodness of fit was evaluated by confirmatory factor analysis. Cronbach's alpha was used to measure the reliability of subscales. Construct validity was tested for different dimensions.ResultsA 9-factor model was obtained, after excluding five “restriction” items and 1 “overt control” item, with a global goodness of fit (CFI = 0.961, TLI = 0.973, RMSEA = 0.057). Cronbach's alpha ranged from 0.684 to 0.889. Children's body mass index (BMI) was significantly and positively related with “perceived parental weight”, “perceived child's weight” and “concern about child's weight”, and inversely related with “pressure to eat”, supporting the theoretical hypothesis. Maternal BMI was positively related with “perceived parental weight”.ConclusionsOur study confirmed the usefulness of this questionnaire for Portuguese preschool children, and supports the need of reformulating the restriction dimension and keeping separately the overt and covert control dimensions.     

Providing personal informal care to older European adults: Should we care about the caregivers' health?

ObjectiveIn Europe, the demand for informal care is high and will increase because of the ageing population. Although caregiving is intended to contribute to the care recipient's health, its effects on the health of older European caregivers are not yet clear. This study explores the association between providing informal personal care and the caregivers' health.MethodData were used from the longitudinal cohort (2004/2005–2010/2011) of the Survey of Health, Ageing and Retirement in Europe (SHARE) (n = 7858). Generalized estimating equations were used to explore the longitudinal association of informal care and the caregiver's health using poor self-rated health (less than good), poor mental health (EURO-D score for depression ≥ 4), and poor physical health (≥ 2 health complaints).ResultsProviding informal personal care was significantly associated with poor mental health (OR = 1.23, 95% CI = 1.04–1.47) and poor physical health (OR = 1.18, 95% CI = 1.01–1.38), after adjusting for various socio-demographic and health-related factors. No statistical significant association was found for self-rated health in the adjusted models.ConclusionProviding informal personal care may negatively influence the caregiver's mental and physical health. More awareness of the beneficial and detrimental effects of caregiving among policy makers is needed to make well-informed decisions concerning the growth of care demands in the ageing population.     

Perception of safety is a prerequisite for the association between neighbourhood green qualities and physical activity: Results from a cross-sectional study in Sweden

In this study, we assess how the Scania Green Score (SGS5), and the five distinct perceived neighbourhood green dimensions within this area-aggregated index (1 km² squares), is associated with self-reported physical activity and general health, and if perceived safety and social coherence has a moderating effect. Two independent surveys, both conducted in Scania, Sweden, was used for data on SGS5 and health outcomes (N=28 198 and N=23 693), respectively. SGS5 was more clearly associated with physical activity (OR 1.06; 95% CI 1.02–1.10) than with general health (OR 1.02; 95% CI 1.00–1.04). This association was moderated by safety (p for interaction <0.001); SGS5 was positively associated with physical activity only among individuals who perceived high safety in their neighbourhood (OR 1.07; 95% CI 1.02–1.11). No moderating effect was seen for social coherence. Among specific dimensions, cultural history was positively associated with both physical activity and general health. Our results suggest that perception of safety is a prerequisite for the positive effects of neighbourhood green qualities.     

Obesity has few effects on future psychosocial functioning of adolescents

We reexamine the effects of obesity on a wide range (n = 17) of indicators of functioning drawn from five broad domains: interpersonal problems, psychological problems, suicidal behaviors, academic performance, and psychiatric disorders. Evidence on this question is mixed. Data are analyzed from a large community sample of adolescents 11–17 at baseline (n = 4175) who were followed up a year later (n = 3134). Using measured height and weight, overweight was defined as 95th > BMI ≤ 85th percentile and obese as BMI > 95th percentile. At baseline, obesity was associated with increased odds only for any mood disorder and poor perceived mental health. For boys, there were no significant associations, but girls had higher odds of problems at school, poor perceived mental health, and mood disorders. Results from the two-wave cohort reveal obesity increased future risk only for poor perceived mental health. For boys, the same pattern was observed, but for girls there were no significant associations. Overall, we found that weight status had few deleterious effects on adolescent social functioning, in multivariate, prospective analyses. If there is an effect of obesity on functioning, it may operate through mediators such as body image.     

Facteurs sociodémographiques et recours aux soins pour raisons psychologiques chez les jeunes adultes

BackgroundAlthough young adults exhibit a high rate of psychiatric disorders, their rate of access to mental health care is low compared with older age groups. Our study examined the relationship between socio-demographic factors and the use of health care services for psychological reasons.MethodsWe studied a community sample of 1103 French 22 to 35-year-old (TEMPO cohort study) who were surveyed by mailed questionnaire in 2009. Data were collected regarding participants’ health (internalizing and externalizing psychological symptoms in 1991 and 2009), health care use (access to health professionals and psychotropic medications in case of psychological difficulties), and socio-demographic factors (sex, age, employment status, marital situation, social support). Parental history of depression was ascertained based on TEMPO participants’ and their parents’ reports (in the GAZEL cohort study).ResultsIn the 12 months preceding the study, 16.7% of study participants saw a health professional and 12.8% took a psychotropic medication for psychological reasons. In multivariate regression, models adjusted for all socio-demographic and psychological characteristics, access to health professionals was associated with being unemployed/out of the labor force (OR = 1.93; 95% CI = 1.11–3.30), family situation (OR in participants living with a partner with no children: 2.16; 95% CI 1.26–3.72; OR in participants not living with a partner: 2.29; 95% CI = 1.34–3.90), and having low social support (OR = 1.75; 95% CI = 1.21–2.54). The use of psychotropic medications was associated with female gender (OR = 2.70; 95% CI = 1.60–4.55), being unemployed/out of the labor force (OR = 3.85; 95% CI = 2.14–6.95), not living with a partner (OR = 2.04; 95% CI = 1.09–3.80) and having low social support (OR = 1.65; 95% CI = 1.05–2.59). Additionally, use of health services was associated with participants’ and their parents’ psychological difficulties.     

Relationship between neighbourhood socioeconomic position and neighbourhood public green space availability: An environmental inequality analysis in a large German city applying generalized linear models

Background

The environmental justice framework states that besides environmental burdens also resources may be social unequally distributed both on the individual and on the neighbourhood level. This ecological study investigated whether neighbourhood socioeconomic position (SEP) was associated with neighbourhood public green space availability in a large German city with more than 1 million inhabitants.

Santé et instabilité professionnelle : résultats issus des centres d’examens de santé de l’assurance-maladie

BackgroundJob insecurity has increased over the last 30 years. Socioeconomic changes have led to various insecure employment categories, including fixed term employment, part time employment and government sponsored jobs. This study was aimed at investigating relationships between employment status and health.MethodsThe study population was composed of 767 184 people, aged 26 to 59 years, examined between 2003 and 2005 in the Health Examination Centers of the French General Health Insurance. Employment status was defined using insecure employment (combining permanent/fixed-term contracts and part-time/full time), government sponsored jobs and duration of unemployment (from < 6 months to ≥ 3 years). Health indicators were poor perceived health, smoking, lack of gynecological follow-up, obesity, untreated caries and high blood pressure. Data were analysed by logistic regression (odds ratios [OR]) adjusted on age, occupational social class and education level, the reference category being permanent full time contracts (OR = 1).ResultsSignificant level-dependent relations with health were observed between non permanent versus permanent employments, part time versus full time. Most OR of unemployed people were higher than those of having employment and increased with duration of unemployment. For example, for poor perceived health in men, OR ranged between 1.00 and 1.68 (95%CI 1.57–1.78) according to job insecurity categories, and between 1.75 (95%CI 1.67–1.83) and 2.80 (95%CI 2.72–2.89) according to duration of unemployment. For obesity in women, OR increased from 1.00 to 1.48 (95%CI 1.37–1.60) in active women and from 1.35 (95%CI 1.27–1.44) to 1.77 (95%CI 1.70–1.84) in unemployed.ConclusionsThis study showed quantitative relationships between job insecurity, unemployment and health. In particular, workers having government-sponsored jobs and long-time unemployed people were at high risk of health problems.