Search

Assessing heterogeneous treatment effects (HTEs) is an essential task in epidemiology. The recent integration of machine learning into causal inference has provided a new, flexible tool for evaluating complex HTEs: causal forest. In a recent paper, Jawadekar et al (Am J Epidemiol. 2023;192(7):1155-1165) introduced this innovative approach and offered practical guidelines for applied users. Building on their work, this commentary provides additional insights and guidance to promote the understanding and application of causal forest in epidemiologic research. We start with conceptual clarifications, differentiating between honesty and cross-fitting, and exploring the interpretation of estimated conditional average treatment effects. We then delve into practical considerations not addressed by Jawadekar et al, including motivations for estimating HTEs, calibration approaches, and ways to leverage causal forest output with examples from simulated data. We conclude by outlining challenges to consider for future advancements and applications of causal forest in epidemiologic research. [ABSTRACT FROM AUTHOR]

In this paper, we present a practical approach for computing the sandwich variance estimator in 2-stage regression model settings. As a motivating example for 2-stage regression, we consider regression calibration, a popular approach for addressing covariate measurement error. The sandwich variance approach has rarely been applied in regression calibration, despite its requiring less computation time than popular resampling approaches for variance estimation, specifically the bootstrap. This is probably because it requires specialized statistical coding. Here we first outline the steps needed to compute the sandwich variance estimator. We then develop a convenient method of computation in R for sandwich variance estimation, which leverages standard regression model outputs and existing R functions and can be applied in the case of a simple random sample or complex survey design. We use a simulation study to compare the sandwich estimator to a resampling variance approach for both settings. Finally, we further compare these 2 variance estimation approaches in data examples from the Women's Health Initiative (1993–2005) and the Hispanic Community Health Study/Study of Latinos (2008–2011). In our simulations, the sandwich variance estimator typically had good numerical performance, but simple Wald bootstrap confidence intervals were unstable or overcovered in certain settings, particularly when there was high correlation between covariates or large measurement error. [ABSTRACT FROM AUTHOR]

"Fusion" study designs combine data from different sources to answer questions that could not be answered (as well) by subsets of the data. Studies that augment main study data with validation data, as in measurement-error correction studies or generalizability studies, are examples of fusion designs. Fusion estimators, here solutions to stacked estimating functions, produce consistent answers to identified research questions using data from fusion designs. In this paper, we describe a pair of examples of fusion designs and estimators, one where we generalize a proportion to a target population and one where we correct measurement error in a proportion. For each case, we present an example motivated by human immunodeficiency virus research and summarize results from simulation studies. Simulations demonstrate that the fusion estimators provide approximately unbiased results with appropriate 95% confidence interval coverage. Fusion estimators can be used to appropriately combine data in answering important questions that benefit from multiple sources of information. [ABSTRACT FROM AUTHOR]

Epidemiologists often compare the observed number of deaths in a cohort with the expected number of deaths, obtained by multiplying person-time accrued in the cohort by mortality rates for a reference population (ideally, a reference that represents the mortality rate in the cohort in the absence of exposure). However, if exposure is hazardous (or salutary), this calculation will not consistently estimate the number of deaths expected in the absence of exposure because exposure will have affected the distribution of person-time observed in the study cohort. While problems with interpretation of this standard calculation of expected counts were discussed more than 2 decades ago, these discussions had little impact on epidemiologic practice. The logic of counterfactuals may help clarify this topic as we revisit these issues. In this paper, we describe a simple way to consistently estimate the expected number of deaths in such settings, and we illustrate the approach using data from a cohort study of mortality among underground miners. [ABSTRACT FROM AUTHOR]

In cohort mortality studies, there often is interest in associations between an exposure of primary interest and mortality due to a range of different causes. A standard approach to such analyses involves fitting a separate regression model for each type of outcome. However, the statistical precision of some estimated associations may be poor because of sparse data. In this paper, we describe a hierarchical regression model for estimation of parameters describing outcome-specific relative rate functions and associated credible intervals. The proposed model uses background stratification to provide flexible control for the outcome-specific associations of potential confounders, and it employs a hierarchical “shrinkage” approach to stabilize estimates of an exposure’s associations with mortality due to different causes of death. The approach is illustrated in analyses of cancer mortality in 2 cohorts: a cohort of dioxin-exposed US chemical workers and a cohort of radiation-exposed Japanese atomic bomb survivors. Compared with standard regression estimates of associations, hierarchical regression yielded estimates with improved precision that tended to have less extreme values. The hierarchical regression approach also allowed the fitting of models with effect-measure modification. The proposed hierarchical approach can yield estimates of association that are more precise than conventional estimates when one wishes to estimate associations with multiple outcomes. [ABSTRACT FROM AUTHOR]

As pellagra reached epidemic proportions in the United States in the early 20th century, 2 teams of investigators assessed its incidence in cotton mill villages in South Carolina. The first, the Thompson-McFadden Commission, concluded that pellagra was likely infectious. The second, a Public Health Service investigation led by Joseph Goldberger, concluded that pellagra was caused by a dietary deficiency. In this paper, we recount the history of the 2 investigations and consider how the differences between the 2 studies' designs, measurements, analyses, and interpretations led to different conclusions. Because the novel dietary assessment strategy was a key feature of the Public Health Service's study design, we incorporated simulated measurement error in a reanalysis of the Public Health Service's data to assess whether this specific difference affected the divergent conclusions. [ABSTRACT FROM PUBLISHER]

The article discusses the binary outcome formula, using counterfactual framework, to account for exposure-mediator interaction in mediation analysis. Topics discussed include effects on longitudinal settings in the presence of time-varying exposure, the marginal structural model approach for continuous outcomes to the setting of binary formula, and the effect of the mediator on the outcome. Also discussed is the effect of the exposure on the mediator.

A propensity score (PS) model's ability to control confounding can be assessed by evaluating covariate balance across exposure groups after PS adjustment. The optimal strategy for evaluating a disease risk score (DRS) model's ability to control confounding is less clear. DRS models cannot be evaluated through balance checks within the full population, and they are usually assessed through prediction diagnostics and goodness-of-fit tests. A proposed alternative is the "dry-run" analysis, which divides the unexposed population into "pseudo-exposed" and "pseudo-unexposed" groups so that differences on observed covariates resemble differences between the actual exposed and unexposed populations. With no exposure effect separating the pseudo-exposed and pseudounexposed groups, a DRS model is evaluated by its ability to retrieve an unconfounded null estimate after adjustment in this pseudo-population. We used simulations and an empirical example to compare traditional DRS performance metrics with the dry-run validation. In simulations, the dry run often improved assessment of confounding control, compared with the C statistic and goodness-of-fit tests. In the empirical example, PS and DRS matching gave similar results and showed good performance in terms of covariate balance (PS matching) and controlling confounding in the dry-run analysis (DRS matching). The dry-run analysis may prove useful in evaluating confounding control through DRS models. [ABSTRACT FROM AUTHOR]

Limitations of extant research on neighborhood disadvantage and health include general reliance on point-intime neighborhood measures and sensitivity to residential self-selection. Using data from the US Census and the 1995-2008 National Longitudinal Study of Adolescent to Adult Health, we applied conventional methods and coarsened exact matching to assess how cardiometabolic health varies among those entering, exiting, or remaining in poor and nonpoor neighborhoods. Within the full sample (n = 11,767), we found significantly higher systolic and diastolic blood pressures among those who entered or consistently lived in poor neighborhoods relative to those who never lived in poor neighborhoods. Obesity was similarly more common among those who originated from poor neighborhoods than among those who originated from nonpoor neighborhoods. Having exited poor neighborhoods was associated with lower systolic blood pressure than was consistent residence in low-income communities. Among the matched sample (n = 9,727), results adjusted for confounders and residential selfselection revealed fewer significant contrasts. Compared with peers who had no neighborhood poverty exposure, those who consistently lived in poor neighborhoods had 46% and 52% higher odds of being obese or hypertensive, respectively. Those who exited neighborhood poverty had significantly higher diastolic blood pressures than those who had never lived in poor neighborhoods. These findings underscore the importance of past as well as current residential circumstances for cardiometabolic health. [ABSTRACT FROM AUTHOR]

Recent studies finding that small-for-gestational-age (SGA) birth is associated with increased adiposity in childhood and adulthood have been based on analyses "adjusting" for height, weight, or body mass index (BMI; weight (kg)/height (m)²) measured concurrently with the adiposity measurement. To assess the potential for bias due to overadjustment for a causal mediator, we compared 2 approaches to analyzing the association between SGA birth and adiposity outcomes (skinfold thicknesses and bioelectrical impedance measurement of body fat) at age 11.5 years using the same data set in a cohort of Belarusian children followed from birth in 1996-1997 to age 11.5 years in 2008-2010. We 1) studied the association of SGA birth with adiposity, adjusting for baseline covariates only, and 2) made additional regression adjustment for concurrent height, weight, or BMI. The first approach yielded negative associations between SGA birth and all adiposity outcomes. Additional adjustment for concurrent weight or BMI reversed (i.e., to positive) the SGA-adiposity association. To explore the latter anthropometric measures as causal mediators, we also used marginal structural models to estimate the controlled direct effect of SGA birth. That effect was similar to the effect seen with the first approach when modeled on height, was null when modeled on BMI, but was confounded by differences in lean mass versus fat mass when modeled on weight. [ABSTRACT FROM AUTHOR]

We evaluated alternative approaches to assessing and correcting for nonresponse bias in a longitudinal survey. We considered the changes in substance-use outcomes over a 3-year period among young adults aged 18-24 years (n = 5,199) in the United States, analyzing data from the National Epidemiologic Survey on Alcohol and Related Conditions. This survey collected a variety of substance-use information from a nationally representative sample of US adults in 2 waves: 2001-2002 and 2004-2005. We first considered nonresponse rates in the second wave as a function of key substance-use outcomes in wave 1. We then evaluated 5 alternative approaches designed to correct for nonresponse bias under different attrition mechanisms, including weighting adjustments, multiple imputation, selection models, and pattern-mixture models. Nonignorable attrition in a longitudinal survey can lead to bias in estimates of change in certain health behaviors over time, and only selected procedures enable analysts to assess the sensitivity of their inferences to different assumptions about the extent of nonignorability. We compared estimates based on these 5 approaches, and we suggest a road map for assessing the risk of nonresponse bias in longitudinal studies. We conclude with directions for future research in this area given the results of our evaluations. [ABSTRACT FROM AUTHOR]

The authors evaluated the validity of a 152-item semiquantitative food frequency questionnaire (SFFQ) by comparing it with two 7-day dietary records (7DDRs) or up to 4 automated self-administered 24-hour recalls (ASA24s) over a 1-year period in the women's Lifestyle Validation Study (2010-2012), conducted among subgroups of the Nurses' Health Studies. Intakes of energy and 44 nutrients were assessed using the 3 methods among 632 US women. Compared with the 7DDRs, SFFQ responses tended to underestimate sodium intake but overestimate intakes of energy, macronutrients, and several nutrients in fruits and vegetables, such as carotenoids. Spearman correlation coefficients between energy-adjusted intakes from 7DDRs and the SFFQ completed at the end of the data-collection period ranged from 0.36 for lauric acid to 0.77 for alcohol (mean r = 0.53). Correlations of the end-period SFFQ were weaker when ASA24s were used as the comparison method (mean r = 0.43). After adjustment for within-person variation in the comparison method, the correlations of the final SFFQ were similar with 7DDRs (mean r = 0.63) and ASA24s (mean r = 0.62). These data indicate that this SFFQ provided reasonably valid estimates for intakes of a wide variety of dietary variables and that use of multiple 24-hour recalls or 7DDRs as a comparison method provided similar conclusions if day-to-day variation was taken into account. [ABSTRACT FROM AUTHOR]

In time-to-event analyses of observational studies of drug effectiveness, incorrect handling of the period between cohort entry and first treatment exposure during follow-up may result in immortal time bias. This bias can be eliminated by acknowledging a change in treatment exposure status with time-dependent analyses, such as fitting a time-dependent Cox model. The prescription time-distribution matching (PTDM) method has been proposed as a simpler approach for controlling immortal time bias. Using simulation studies and theoretical quantification of bias, we compared the performance of the PTDM approach with that of the time-dependent Cox model in the presence of immortal time. Both assessments revealed that the PTDM approach did not adequately address immortal time bias. Based on our simulation results, another recently proposed observational data analysis technique, the sequential Cox approach, was found to be more useful than the PTDM approach (Cox: bias = -0.002, mean squared error = 0.025; PTDM: bias = -1.411, mean squared error = 2.011). We applied these approaches to investigate the association of β-interferon treatment with delaying disability progression in a multiple sclerosis cohort in British Columbia, Canada (Long-Term Benefits and Adverse Effects of Beta-Interferon for Multiple Sclerosis (BeAMS) Study, 1995-2008). [ABSTRACT FROM AUTHOR]

The growing prevalence of overweight and obesity among children is well documented, but prevalence estimates offer little insight into rates of transition to higher or lower body mass index (BMI; weight (kg)/height (m)²) categories. We estimated the expected numbers of years children would live as normal weight, overweight, and obese by race/ ethnicity and sex, given rates of transition across BMI status levels. We used multistate life table methods and transition rates estimated from prospective cohort data (2007-2013) for Denver, Colorado, public schoolchildren aged 3-15 years. At age 3 years, normal-weight children could expect to live 11.1 of the following 13 years with normal weight status, and obese children could expect to live 9.8 years with obese status. At age 3 years, overweight children could expect to live 4.5 of the following 13 years with normal weight status, 5.1 years with overweight status, and 3.4 years with obese status. Whites and Asians lived more years at lower BMI status levels than did blacks or Hispanics; sex differences varied by race/ethnicity. Children who were normal weight or obese at age 3 years were relatively unlikely to move into a different BMI category by age 15 years. Overweight children are relatively likely to transition to normal weight or obese status. [ABSTRACT FROM AUTHOR]

This study's objective was to investigate healthy aging in older French adults 5 years after a period of daily nutritional-dose supplementation with antioxidant nutrients. The study was based on the double-blind, randomized trial, Supplementation with Antioxidant Vitamins and Minerals (SU.VI.MAX) Study (1994-2002) and the SU.VI.MAX 2 Follow-up Study (2007-2009). During 1994-2002, participants received a daily combination of vitamin C (120 mg), β-carotene (6 mg), vitamin E (30 mg), selenium (100 µg), and zinc (20 mg) or placebo. Healthy aging was assessed in 2007-2009 by using multiple criteria, including the absence of major chronic disease and good physical and cognitive functioning. Data from a subsample of the SU.VI.MAX 2 cohort, initially free of major chronic disease, with a mean age of 65.3 years in 2007-2009 (n = 3,966), were used to calculate relative risks. Supplementation was associated with a greater healthy aging probability among men (relative risk = 1.16, 95% confidence interval: 1.04, 1.29) but not among women (relative risk = 0.98, 95%confidence interval: 0.86, 1.11) or all participants (relative risk = 1.07, 95%confidence interval: 0.99, 1.16). Moreover, exploratory subgroup analyses indicated effect modification by initial serum concentrations of zinc and vitamin C. In conclusion, an adequate supply of antioxidant nutrients (equivalent to quantities provided by a balanced diet rich in fruits and vegetables) may have a beneficial role for healthy aging. [ABSTRACT FROM AUTHOR]

To lower the prevalence of hypertension and racial disparities in hypertension, public health agencies have attempted to reduce modifiable risk factors for high blood pressure, such as excess sodium intake or high body mass index. In the present study, we used decomposition methods to identify how population-level reductions in key risk factors for hypertension could reshape entire population distributions of blood pressure and associated disparities among racial/ethnic groups. We compared blood pressure distributions among non-Hispanic white, non-Hispanic black, and Mexican-American persons using data from the US National Health and Nutrition Examination Survey (2003-2010). When using standard adjusted logistic regression analysis, we found that differences in body mass index were the only significant explanatory correlate to racial disparities in blood pressure. By contrast, our decomposition approach provided more nuanced revelations; we found that disparities in hypertension related to tobacco use might be masked by differences in body mass index that significantly increase the disparities between black and white participants. Analysis of disparities between white and Mexican-American participants also reveal hidden relationships between tobacco use, body mass index, and blood pressure. Decomposition offers an approach to understand how modifying risk factors might alter population-level health disparities in overall outcome distributions that can be obscured by standard regression analyses. [ABSTRACT FROM AUTHOR]

Ungrouping binned data can be desirable for many reasons: Bins can be too coarse to allow for accurate analysis; comparisons can be hindered when different grouping approaches are used in different histograms; and the last interval is often wide and open-ended and, thus, covers a lot of information in the tail area. Age group-specific disease incidence rates and abridged life tables are examples of binned data. We propose a versatile method for ungrouping histograms that assumes that only the underlying distribution is smooth. Because of this modest assumption, the approach is suitable for most applications. The method is based on the composite link model, with a penalty added to ensure the smoothness of the target distribution. Estimates are obtained by maximizing a penalized likelihood. This maximization is performed efficiently by a version of the iteratively reweighted least-squares algorithm. Optimal values of the smoothing parameter are chosen by minimizing Akaike’s Information Criterion.We demonstrate the performance of this method in a simulation study and provide several examples that illustrate the approach. Wide, open-ended intervals can be handled properly. The method can be extended to the estimation of rates when both the event counts and the exposures to risk are grouped. [ABSTRACT FROM AUTHOR]

Urban form—the structure of the built environment—can influence physical activity, yet little is known about how walkable design differs according to neighborhood sociodemographic composition. We studied how walkable urban form varies by neighborhood sociodemographic composition, region, and urbanicity across the United States. Using linear regression models and 2000–2001 US Census data, we investigated the relationship between 5 neighborhood census characteristics (income, education, racial/ethnic composition, age distribution, and sex) and 5 walkability indicators in almost 65,000 census tracts in 48 states and the District of Columbia. Data on the built environment were obtained from the RAND Corporation's (Santa Monica, California) Center for Population Health and Health Disparities (median block length, street segment, and node density) and the US Geological Survey's National Land Cover Database (proportion open space and proportion highly developed). Disadvantaged neighborhoods and those with more educated residents were more walkable (i.e., shorter block length, greater street node density, more developed land use, and higher density of street segments). However, tracts with a higher proportion of children and older adults were less walkable (fewer street nodes and lower density of street segments), after adjustment for region and level of urbanicity. Research and policy on the walkability-health link should give nuanced attention to the gap between persons living in walkable areas and those for whom walkability has the most to offer. [ABSTRACT FROM PUBLISHER]

The C statistic is a commonly reported measure of screening test performance. Optimistic estimation of the C statistic is a frequent problem because of overfitting of statistical models in small data sets, and methods exist to correct for this issue. However, many studies do not use such methods, and those that do correct for optimism use diverse methods, some of which are known to be biased. We used clinical data sets (United Kingdom Down syndrome screening data from Glasgow (1991–2003), Edinburgh (1999–2003), and Cambridge (1990–2006), as well as Scottish national pregnancy discharge data (2004–2007)) to evaluate different approaches to adjustment for optimism. We found that sample splitting, cross-validation without replication, and leave-1-out cross-validation produced optimism-adjusted estimates of the C statistic that were biased and/or associated with greater absolute error than other available methods. Cross-validation with replication, bootstrapping, and a new method (leave-pair-out cross-validation) all generated unbiased optimism-adjusted estimates of the C statistic and had similar absolute errors in the clinical data set. Larger simulation studies confirmed that all 3 methods performed similarly with 10 or more events per variable, or when the C statistic was 0.9 or greater. However, with lower events per variable or lower C statistics, bootstrapping tended to be optimistic but with lower absolute and mean squared errors than both methods of cross-validation. [ABSTRACT FROM PUBLISHER]

There has been a worldwide epidemic of obesity in recent decades. In animal studies, there is convincing evidence that light exposure causes weight gain, even when calorie intake and physical activity are held constant. Disruption of sleep and circadian rhythms by exposure to light at night (LAN) might be one mechanism contributing to the rise in obesity, but it has not been well-investigated in humans. Using multinomial logistic regression, we examined the association between exposure to LAN and obesity in questionnaire data from over 100,000 women in the Breakthrough Generations Study, a cohort study of women aged 16 years or older who were living in the United Kingdom and recruited during 2003–2012. The odds of obesity, measured using body mass index, waist:hip ratio, waist:height ratio, and waist circumference, increased with increasing levels of LAN exposure (P < 0.001), even after adjustment for potential confounders such as sleep duration, alcohol intake, physical activity, and current smoking. We found a significant association between LAN exposure and obesity which was not explained by potential confounders we could measure. While the possibility of residual confounding cannot be excluded, the pattern is intriguing, accords with the results of animal experiments, and warrants further investigation. [ABSTRACT FROM PUBLISHER]