Epidemiology最新文献

Background: Childhood maltreatment is associated with elevated adult weight. It is unclear whether this association extends to pregnancy, a critical window for the development of obesity.

Methods: We examined associations of childhood maltreatment histories with pre-pregnancy BMI and gestational weight gain among women who had participated for >20 years in a longitudinal cohort.At age 26-35 participants reported childhood maltreatment (physical, sexual, and emotional abuse; emotional neglect) and, 5 years later, about pre-pregnancy weight and gestational weight gain for previous pregnancies (n=656). Modified Poisson regression models were used to estimate associations of maltreatment history with pre-pregnancy BMI and gestational weight gain z-scores, adjusting for sociodemographics. We used Multivariate Imputation by Chained Equations to adjust outcome measures for misclassification using data from an internal validation study.

Results: Before misclassification adjustment, results indicated a higher risk of pre-pregnancy BMI ≥30 kg/m2 in women with certain types of maltreatment (e.g., emotional abuse RR=2.4; 95% CI: 1.5, 3.7) compared with women without that maltreatment type. After misclassification adjustment, estimates were attenuated but still modestly elevated (e.g., emotional abuse RR=1.7; 95% CI: 1.1, 2.7). Misclassification-adjusted estimates for maltreatment associations with gestational weight gain z-scores were close to the null and imprecise.

Conclusions: Findings suggest an association of maltreatment with pre-pregnancy BMI ≥30 kg/m2 but not with high gestational weight gain. Results suggest a potential need for equitable interventions that can support all women, including those with maltreatment histories, as they enter pregnancy.

Machine learning techniques for causal effect estimation can enhance the reliability of epidemiologic analyses, reducing their dependence on correct model specifications. However, the stochastic nature of many machine learning algorithms implies that the results derived from such approaches may be influenced by the random seed that is set prior to model fitting. In this work, we highlight the substantial influence of random seeds on a popular approach for machine learning-based causal effect estimation, namely doubly robust estimators. We illustrate that varying seeds can yield divergent scientific interpretations of doubly robust estimates produced from the same dataset. We propose techniques for stabilizing results across random seeds and, through an extensive simulation study, demonstrate that these techniques effectively neutralize seed-related variability without compromising the statistical efficiency of the estimators. Based on these findings, we offer practical guidelines to minimize the influence of random seeds in real-world applications, and we encourage researchers to explore variability due to random seed when implementing any method that involves random steps.

Background: Use of machine learning to estimate exposure effects introduces a dependence between the results of an empirical study and the value of the seed used to fix the pseudo-random number generator.

Methods: We used data from 10,038 pregnant women and a 10% subsample (N = 1,004) to examine the extent to which the risk difference for the relation between fruit and vegetable consumption and preeclampsia risk changes under different seed values. We fit an augmented inverse probability weighted estimator with two Super Learner algorithms: a simple algorithm including random forests and single layer neural networks and a more complex algorithm with a mix of tree-based, regression based, penalized and simple algorithms. We evaluated the distributions of risk differences, standard errors, and p values that result from 5,000 different seed value selections.

Results: Our findings suggest important variability in the risk difference estimates, as well as an important effect of the stacking algorithm used. The interquartile range width (IQRw) of the risk differences in the full sample with the simple algorithm was 13 per 1000. However, all other IQRs were roughly an order of magnitude lower. The medians of the distributions of risk differences differed according to the sample size and the algorithm used.

Conclusions: Our findings add another dimension of concern regarding the potential for "p-hacking", and further warrants the need to move away from simplistic evidentiary thresholds in empirical research. When empirical results depend on pseudo-random number generator seed values, caution is warranted in interpreting these results.

Surveys are commonly used to facilitate research in epidemiology, health, and the social and behavioral sciences. Often, these surveys are not simple random samples, and respondents are given weights reflecting their probability of selection into the survey. We show that using survey weights can be beneficial for evaluating the quality of predictive models when splitting data into training and test sets. In particular, we characterize model assessment statistics, such as sensitivity and specificity, as finite population quantities and compute survey-weighted estimates of these quantities with test data comprising a random subset of the original data. Using simulations with data from the National Survey on Drug Use and Health and the National Comorbidity Survey, we show that unweighted metrics estimated with sample test data can misrepresent population performance, but weighted metrics appropriately adjust for the complex sampling design. We also show that this conclusion holds for models trained using upsampling for mitigating class imbalance. The results suggest that weighted metrics should be used when evaluating performance on test data derived from complex surveys.

Most prior work in quantitative approaches to generalizability and transportability emphasizes extending causal effect estimates from randomized trials to target populations. Extending findings from observational studies is also of scientific interest, and identifiability assumptions and estimation methods differ from randomized settings when there is selection on both the exposure and exposure-outcome mediators in combination with exposure-outcome confounders (and both confounders and mediators can modify exposure-outcome effects). We argue that this causal structure is common in observational studies, particularly in the field of lifecourse epidemiology, e.g., when extending estimates of the effect of an early-life exposure on a later-life outcome from a cohort enrolled in mid- to late-life. We describe identifiability assumptions and identification using observed data in such settings, highlighting differences from work extending findings from randomized trials. We describe statistical methods, including weighting, outcome modeling, and doubly robust approaches to estimate potential outcome means and verage treatment effects in the target population and illustrate performance of the methods in a simulation study. We show that in the presence of selection into the study sample on both exposure and confounders, estimators must be able to address confounding in the target population. When there is also selection on mediators of the exposure-outcome relationship, estimators need to be able to use different sets of variables to account for selection (including the mediator), and confounding. We discuss conceptual implications of our results, as well as highlight unresolved practical questions for applied work to extend findings from observational studies to target populations.

Background: Extreme ambient heat is unambiguously associated with higher risk of illness and death. The Optum Labs Data Warehouse (OLDW), a database of medical claims from US-based patients with commercial or Medicare Advantage health insurance, has been used to quantify heat-related health impacts. Whether results for the insured sub-population are generalizable to the broader population has to our knowledge not been documented. We sought to address this question, for the US population in California from 2012 to 2019.

Methods: We examined changes in daily rates of emergency department (ED) encounters and in-patient hospitalization encounters for all-causes, heat-related outcomes, renal disease, mental/behavioral disorders, cardiovascular disease, and respiratory disease. OLDW was the source for health data for insured individuals in California, and health data for the broader population were gathered from the California Department of Health Care Access and Information (HCAI). We defined extreme heat exposure as any day in a group of 2 or more days with maximum temperatures exceeding the county-specific 97.5 th percentile and used a space-time-stratified case-crossover design to assess and compare the impacts of heat on health.

Results: Average incidence rates of medical encounters differed by dataset. However, rate ratios for ED encounters were similar across datasets for all causes (ratio of incidence rate ratios (rIRR) = 0.989; 95% confidence interval (CI) = 0.973, 1.011), heat-related causes (rIRR = 1.080; 95% CI = 0.999, 1.168), renal disease (rIRR = 0.963; 95% CI = 0.718, 1.292), and mental health disorders (rIRR = 1.098; 95% CI = 1.004, 1.201). Rate ratios for inpatient encounters were also similar.

Conclusions: This work presents evidence that OLDW can continue to be a resource for estimating the health impacts of extreme heat.