American journal of epidemiology
-
We investigated the association between dairy product consumption and all-cause, cardiovascular disease (CVD), and cancer mortality in the Golestan Cohort Study, a prospective cohort study launched in January 2004 in Golestan Province, northeastern Iran. A total of 42,403 men and women participated in the study and completed a diet questionnaire at enrollment. Cox proportional hazards models were used to estimate hazard ratios and 95% confidence intervals. ⋯ Higher intake of high-fat dairy food and milk was not associated with all-cause or CVD mortality. Neither intake of individual dairy products nor intake of total dairy products was significantly associated with overall cancer mortality. High consumption of dairy products, especially yogurt and cheese, may reduce the risk of overall and CVD mortality.
-
Investigating firearm injury trends over the past decade, we examined temporal trends overall and according to race/ethnicity and intent in fatal and nonfatal firearm injuries (FFIs and NFIs) in United States during 2001-2013. Counts of FFIs and estimated counts of NFIs were obtained from the Centers for Disease Control and Prevention's Web-based Injury Statistics Query and Reporting System. Poisson regression was used to analyze overall and subgroup temporal trends and to estimate annual change per 100,000 persons (change). ⋯ Among whites, FFIs (change = 0.15; Ptrend < 0.0001) and NFIs (change = 0.13; Ptrend < 0.0001) increased; among blacks, FFIs declined (change = -0.20; Ptrend < 0.0001). Among Hispanics, FFIs declined (change = -0.28; Ptrend < 0.0001) while NFIs increased (change = 0.55; Ptrend = 0.014). The endemic firearm-related injury rates during the first decade of the 21st century mask a shift from firearm deaths towards a rapid rise in nonfatal injuries.
-
Aspirin, nonsteroidal antiinflammatory drugs (NSAID), and acetaminophen are commonly used. Frequent use of analgesics has been associated with a higher risk of hearing loss. However, the association between duration of analgesic use and the risk of hearing loss is unclear. ⋯ Duration of aspirin use was not associated with hearing loss (for >6 years of use compared with <1 year, multivariable-adjusted relative risk = 1.01, 95% confidence interval: 0.97, 1.05; P for trend = 0.35). In this cohort of women, longer durations of NSAID and acetaminophen use were associated with slightly higher risks of hearing loss, but duration of aspirin use was not. Considering the high prevalence of analgesic use, this may be an important modifiable contributor to hearing loss.
-
Estimation of causal effects using observational data continues to grow in popularity in the epidemiologic literature. While many applications of causal effect estimation use propensity score methods or G-computation, targeted maximum likelihood estimation (TMLE) is a well-established alternative method with desirable statistical properties. TMLE is a doubly robust maximum-likelihood-based approach that includes a secondary "targeting" step that optimizes the bias-variance tradeoff for the target parameter. ⋯ Our simulation study compares methods under parametric regression misspecification; our results highlight TMLE's property of double robustness. Additionally, we discuss best practices for TMLE implementation, particularly the use of ensembled machine learning algorithms. Our simulation study demonstrates all methods using super learning, highlighting that incorporation of machine learning may outperform parametric regression in observational data settings.
-
The US physical activity (PA) recommendations were based primarily on studies in which self-reported data were used. Studies that include accelerometer-assessed PA and sedentary behavior can contribute to these recommendations. In the present study, we explored the associations of PA and sedentary behavior with all-cause and cardiovascular disease (CVD) mortality in a nationally representative sample. ⋯ Lower all-cause and CVD mortality risks were also generally observed for persons with higher accelerometer-assessed moderate and moderate-to-vigorous PA levels and for self-reported moderate-to-vigorous leisure, household and total activities, as well as for meeting PA recommendations. Accelerometer-assessed sedentary behavior was generally not associated with all-cause or CVD mortality in fully adjusted models. These findings support the national PA recommendations to reduce mortality.