2324

Background: While breast cancer (BC) is associated with long-term survival, especially in early stage cases, factors influencing BC progression and survival are not perfectly understood. Residential exposure to ambient particulate matter <2.5 μg/m³ (PM_2.5) has been associated to increased BC incidence; but only three studies focused on BC survival, with no consistent findings. Objective: To assess the associations between mean annual PM_2.5 exposure after diagnosis, and total and BC specific mortality. Methods: A total of 4,804 women from the French E3N cohort study and diagnosed with BC between 1990 and 2008 were followed through December 2011. Mean residential PM_2.5 exposure was derived from Europe-wide hybrid land use regression models at a 100 m spatial grid (ELAPSE project). Cox regression analysis was used to estimate hazard ratios (HR) and 95% confidence intervals (CI) of total and BC-specific mortality per increase in annual PM_2.5 exposure, controlling for BC risk factors. We also evaluated possible non-linear dose-response relationship by using restricted cubic splines, and investigated potential effect modification by clinical and lifestyle factors. Results: A total of 608 women died over the follow-up (among whom 52% of BC-specific deaths). Mean annual PM_2.5 concentration was 3.6 μg/m³ (inter-quartile range= 19.6 μg/m³). Overall, a 10 μg/m³ increase in PM_2.5 was associated with modest increased total (HR= 1.18; 95% CI: 0.85-1.63) and BC-specific (HR= 1.17; 95% CI: 0.74-1.84) mortality. Associations per 10 μg/m³ increase in PM_2.5 were stronger among stage 1 BC cases [HR= 1.57 (95% CI: 0.95-2.60) and HR= 2.43 (95% CI: 1.17-5.08), respectively] with total and BC-specific mortality, respectively, than in higher stage BC (P_Interaction = 0.05 and 0.02, respectively). There was no evidence of a non-linear relationship. Conclusion: Our findings suggest that exposure to high levels of PM_2.5 could reduce long-term survival after diagnosis of a good-prognosis BC.

2815

Background: Chronic exposure to metals has been associated with adverse neurological outcomes in the elderly. Although the mechanisms are not completely understood, the inflammatory processes in the brain are suspected as a pathway by which metals exerts their neurotoxicity. In parallel, a high diet quality may protect against chronic inflammation. Objective: We sought to examine the joint associations of overall diet quality - as measured by the Adapted Dietary Inflammatory Index (ADII) and the Healthy Eating Index 2015 (HEI-2015) - and blood metals concentrations on cognitive performance in elderly. Methods: We used data on 1,777 adults aged ≥60 years from the US National Health and Nutrition Examination Survey (NHANES; 2011-2014). We derived the ADII and the HEI-2015 from two nonconsecutive 24-hour diet recalls. Cognitive performance was measured by the Consortium to Establish a Registry for Alzheimer’s disease, the Animal Fluency test, and the Digit Symbol Substitution test. We calculated a composite z-score reflecting the overall cognitive performance. The effect of diet quality and blood metals concentrations on cognitive performances were assessed using generalized linear and additive models adjusted for appropriate confounders. Results: Mean score of overall cognitive performance was 87.8 (SD = 24.3). Higher HEI-2015 scores (reflecting higher diet quality), were associated with higher overall cognitive performance [B= 0.09; 95%CI: 0.02-0.16], while higher ADII scores (reflecting a pro-inflammatory diet) with lower cognitive performance [B= -0.09; 95%CI: -0.17;-0.02]. High lead concentrations were associated with poorer overall cognitive performance. However, this was observed only in subjects with low HEI-2015 or high ADII scores, while there was no association with lead concentrations in subjects with better-quality diets (p-interaction <0.08 for all). Conclusion: Our findings suggest that a better-quality diet, as assessed with a high HEI-2015 or a low ADII score may help prevent blood lead adverse cognitive effects among elderly people.