Prevalence of high blood pressure differs across regions in France: Estimations from a cross-sectional analysis of the CONSTANCES cohort

International audienceIntroduction In France, important territorial variations of diseases such as cardiovascular or chronic kidney diseases have been reported. For example, incidence of end stage renal disease in the east of France is twofold higher compared to that in the north. These diseases share common risk factors, mainly high blood pressure (HBP) which is the leading preventable cause of premature death worldwide. However, few studies had the scope and sufficient power to explore territorial variations of HBP within France. The aims of this study are first to identify the geographical variations of prevalence of HBP across the country using the data from the French cohort CONSTANCES; second to explore the potential contributions of behavioral, socioeconomic and access to care characteristics in these variations. Methods CONSTANCES is a population-based cohort targeting adults aged 18 to 69 years old living in different regions of France. Information on lifestyle was collected by self-administered questionnaire. Blood pressure was measured in 16 different recruitment centers using a standardized protocol. HBP has been defined as blood pressure over 140/90 mmHg and/or taking antihypertensive medication as indicated by matched records from the national database or reimbursements from the French health insurance. Analyses were stratified by gender. We first calculated prevalence of HBP as well as proportion of treated hypertensive participants across the different centers. We then performed logistic regressions to estimate changes in predicted prevalence of HBP between centers after taking into account individual behavioral and socioeconomic as well as neighborhood variables. Results A total of 33,665 individuals (52.5% women) recruited in 2013 and 2014 were included in the analyses. Mean age in men and women was 48.8 ± 12.9 years and 47.3 ± 13.0 years respectively. In total, 10,285 individuals were classified hypertensive, yielding a crude prevalence of HBP of 30.6%. Using 2014 French population as reference, age-standardized prevalence of HBP was 27.1% [95% CI: 26.6–27.7], higher among men (33.7% [95% CI: 32.8–34.6%]) than women (20.8% [95% CI: 20.1–21.4%]). Prevalence increased with age, from 10.3% [95% CI: 9.5–11.2] in 18–34 years to 62.7% [95% CI: 60.2–65.2] in people aged 65 years and more. Pronounced differences between regions were observed. The region of Lille had the highest prevalence both in men (48.6% [95% CI: 44.5–52.6%]) and women (28.3% [95% CI: 25.6–31.1%]), and in all age groups. The lowest prevalence was observed in Paris in men (26.0% [95% CI: 24.2–27.9]) and in the region of Saint-Brieuc in women (16.5% [95% CI: 14.5–18.4]). The proportion of treated hypertensive individuals also varied across regions, from 37.1% in the region of Saint-Nazaire to more than 66% in Saint-Brieuc in women and from 32.5% in Lille to 49.1% in Saint-Brieuc in men. Taking into account behavioral risk factors reduced moderately variations in prevalence of HBP. In men, adjusting on body mass index, physical activity and smoking status was associated with a 3.9% > diminution of the difference between extreme regions in prevalence of HBP (from 22.7% to 18.8%). Adjusting in addition on socioeconomic individual factors such as education, accessibility to a general practitioner or a pharmacist, and on rural/urban typology of residence was associated with a further 3.5% > diminution of the difference between extreme regions in prevalence of HBP (from 18.8% to 15.3%). Thus the global reduction from 22.7% to 15.3% corresponds to a decrease of around 33% in the spread of territorial variations. Conclusions Prevalence of HBP differs markedly between geographical areas in France. The gradient observed in this analysis is in accordance with results from the previous French MONICA study, with a higher prevalence of HBP in the north and the east of France compared to the south. Behavioral and socioeconomic risk factors explain nearly a third of the difference in prevalence of HBP between regions