Search

Limited information is available on potential predictive value of environmental chemicals for mortality. Our study aimed to investigate the associations between 43 of 8 classes representative environmental chemicals in serum/urine and mortality, and further develop the interpretable machine learning models associated with environmental chemicals to predict mortality. A total of 1602 participants were included from the National Health and Nutrition Examination Survey (NHANES). During 154,646 person-months of follow-up, 127 deaths occurred. We found that machine learning showed promise in predicting mortality. CoxPH was selected as the optimal model for predicting all-cause mortality with time-dependent AUROC of 0.953 (95%CI: 0.951–0.955). Coxnet was the best model for predicting cardiovascular disease (CVD) and cancer mortality with time-dependent AUROCs of 0.935 (95%CI: 0.933–0.936) and 0.850 (95%CI: 0.844–0.857). Based on clinical variables, adding environmental chemicals could enhance the predictive ability of cancer mortality (P < 0.05). Some environmental chemicals contributed more to the models than traditional clinical variables. Combined the results of association and prediction models by interpretable machine learning analyses, we found urinary methyl paraben (MP) and urinary 2-napthol (2-NAP) were negatively associated with all-cause mortality, while serum cadmium (Cd) was positively associated with all-cause mortality. Urinary bisphenol A (BPA) was positively associated with CVD mortality. • Machine learning showed promise in predicting mortality. • Based on clinical variables, adding environmental chemicals could enhance the predictive ability of cancer mortality. • Serum Cd was positively correlated with all-cause mortality, while negative association was shown for urinary MP and 2-NAP. • Urinary BPA was positively associated with CVD mortality. [ABSTRACT FROM AUTHOR]

Evidence regarding the role of physical frailty in cancer-related outcomes is limited. We aimed to examine the association of frailty with cancer incidence and mortality risk. This prospective study included 348,144 participants free of cancer at baseline from the UK Biobank. Frailty phenotypes (non-frail, pre-frail, and frail) were constructed from 5 components: weight loss, exhaustion, low physical activity, slow gait speed, and low grip strength. The outcome was incidence and mortality of overall and cite-specific cancers. Cox proportional hazard regression was used to estimate the association of frailty phenotypes with cancer incidence and mortality risk. A total of 43,304 incident cancer cases and 10,152 cancer deaths were documented during a median of 12.0 years of follow-up. For overall cancer, compared with non-frailty, the incidence risk increased by 4% for pre-frailty and 11% for frailty, and the mortality risk increased by 11% for pre-frailty and 39% for frailty. Frailty phenotypes were also dose-dependently associated with a higher risk of incidence and mortality of some site-specific cancers (including liver and lung), with significant sex differences. We observed a synergetic association of frailty phenotypes and smoking with overall cancer incidence and mortality risk. Frailty phenotypes contributed significantly to a higher risk of overall and some site-specific cancers incidence and mortality in a stepwise manner or within individual categories. Future studies are warranted to emphasize the identification, management and prevention of frailty in the whole population and complements of lifestyle-targeted interventions such as quitting smoking. • Evidence about the associations of frailty with cancer outcomes are insufficient. • Pre-frailty and frailty contributed to higher overall and site-specific cancer risks. • A synergetic effect between frailty phenotypes and smoking existed. [ABSTRACT FROM AUTHOR]

Objective: The lack of health insurance reduces access to care and often results in poorer health outcomes. The present study simultaneously assessed the effects of health insurance on cancer and chronic disease mortality, as well as the inter-relationships with diet, obesity, smoking, and inflammatory biomarkers. We hypothesized that public/no insurance versus private insurance would result in increased cancer/chronic disease mortality due to the increased prevalence of inflammation-related lifestyle factors in the underinsured population.Methods: Data from the Third National Health and Nutrition Examination Survey participants (NHANES III;1988-1994) were prospectively examined to assess the effects of public/no insurance versus private insurance and inflammation-related lifestyle factors on mortality risk from cancer, all causes, cardiovascular disease (CVD) and diabetes. Cox proportional hazards regression was performed to assess these relationships.Results: Multivariate regression analyses revealed substantially greater risks of mortality ranging from 35% to 245% for public/no insurance versus private insurance for cancer (HR=1.35; 95% CI=1.09,1.66), all causes (HR=1.54; 95% CI=1.39,1.70), CVD (HR=1.62; 95% CI=1.38,1.90) and diabetes (HR=2.45; 95% CI=1.45,4.14). Elevated CRP, smoking, reduced diet quality and higher BMI were more prevalent in those with public insurance, and were also associated with increased risks of cancer/chronic disease mortality.Discussion: Insurance status was strongly associated with cancer/chronic disease mortality after adjusting for lifestyle factors. The results suggest that inadequate health insurance coverage results in a substantially greater need for preventive strategies that focus on tobacco control, obesity, and improved dietary quality. These efforts should be incorporated into comprehensive insurance coverage programs for all Americans. [ABSTRACT FROM AUTHOR]

The aim was to investigate associations between health locus of control (HLC) and all-cause, cardiovascular (CVD), cancer and other cause mortality. A public health postal questionnaire was distributed in the autumn of 2008 to a stratified random sample of the 18-80 year old adult population in Scania in southernmost Sweden. The participation rate was 54.1%, and 25,517 participants were included in the present study. Baseline 2008 survey data was linked to cause of death register data to create a prospective cohort with 8.3-year follow-up. Associations between health locus of control and mortality were investigated in survival (Cox) regression models. Prevalence of internal HLC was 69.0% and external HLC 31.0% among women. Internal HLC was 67.6% and external HLC 32.4% among men. In the models with women and men combined, external HLC had significantly higher all-cause, CVD, cancer and other cause mortality even after adjustments for sociodemographic factors and chronic disease at baseline, but after the introduction of health-related behaviors, external HLC only displayed higher cancer mortality compared to internal HLC. External HLC displayed higher all-cause, cancer and other cause mortality for men in the final model adjusted for health-related behaviors, but not for women. Other pathways than health-related behaviors may exist for the association between external HLC and cancer mortality, particularly among men. [ABSTRACT FROM AUTHOR]