Utility of Wrist-Wearable Data for Assessing Pain, Sleep, and Anxiety Outcomes After Traumatic Stress Exposure.

Importance: Adverse posttraumatic neuropsychiatric sequelae after traumatic stress exposure are common and have higher incidence among socioeconomically disadvantaged populations. Pain, depression, avoidance of trauma reminders, reexperiencing trauma, anxiety, hyperarousal, sleep disruption, and nightmares have been reported. Wrist-wearable devices with accelerometers capable of assessing 24-hour rest-activity characteristics are prevalent and may have utility in measuring these outcomes. Objective: To evaluate whether wrist-wearable devices can provide useful biomarkers for recovery after traumatic stress exposure. Design, Setting, and Participants: Data were analyzed from a diverse cohort of individuals seen in the emergency department after experiencing a traumatic stress exposure, as part of the Advancing Understanding of Recovery After Trauma (AURORA) study. Participants recruited from 27 emergency departments wore wrist-wearable devices for 8 weeks, beginning in the emergency department, and completed serial assessments of neuropsychiatric symptoms. A total of 19 019 patients were screened. Of these, 3040 patients met study criteria, provided informed consent, and completed baseline assessments. A total of 2021 provided data from wrist-wearable devices, completed the 8-week assessment, and were included in this analysis. The data were randomly divided into 2 equal parts (n = 1010) for biomarker identification and validation. Data were collected from September 2017 to January 2020, and data were analyzed from May 2020 to November 2022. Exposures: Participants were recruited for the study after experiencing a traumatic stress exposure (most commonly motor vehicle collision). Main Outcomes and Measures: Rest-activity characteristics were derived and validated from wrist-wearable devices associated with specific self-reported symptom domains at a point in time and changes in symptom severity over time. Results: Of 2021 included patients, 1257 (62.2%) were female, and the mean (SD) age was 35.8 (13.0) years. Eight wrist-wearable device biomarkers for symptoms of adverse posttraumatic neuropsychiatric sequelae exceeded significance thresholds in the derivation cohort. One of these, reduced 24-hour activity variance, was associated with greater pain severity (r = -0.14; 95% CI, -0.20 to -0.07). Changes in 6 rest-activity measures were associated with changes in pain over time, and changes in the number of transitions between sleep and wake over time were associated with changes in pain, sleep, and anxiety. Simple cutoffs for these biomarkers identified individuals with good recovery for pain (positive predictive value [PPV], 0.85; 95% CI, 0.82-0.88), sleep (PPV, 0.63; 95% CI, 0.59-0.67, and anxiety (PPV, 0.76; 95% CI, 0.72-0.80) with high predictive value. Conclusions and Relevance: These findings suggest that wrist-wearable device biomarkers may have utility as screening tools for pain, sleep, and anxiety symptom outcomes after trauma exposure in high-risk populations.

Comparing estimators for latent interaction models under structural and distributional misspecifications.

Estimation methods for structural equation models with interactions of latent variables were compared in several studies. Yet none of these studies examined models that were structurally misspecified. Here, the model-implied instrumental variable 2-stage least square estimator (MIIV-2SLS; Bollen, 1995; Bollen & Paxton, 1998), the 2-stage method of moments estimator (2SMM; Wall & Amemiya, 2003), the nonlinear structural equation mixture model approach (NSEMM; Kelava, Nagengast, & Brandt, 2014), and the unconstrained product indicator approach (UPI; Marsh, Wen, & Hau, 2004) were compared in a Monte Carlo simulation. The design included structural misspecifications in the measurement model involving the scaling indicator or not, the size of the misspecification, normal and nonnormal data, the indicators' reliability, and sample size. For the structural misspecifications that did not involve the scaling indicator, we found that MIIV-2SLS' parameter estimates were less biased compared with 2SMM, NSEMM, and UPI. If the reliability was high, the RMSE for all approaches was very similar; for low reliability, MIIV-2SLS' RMSE became larger compared with the other approaches. If the structural misspecification involved the scaling indicator, all estimators were seriously biased, with the largest bias for MIIV-2SLS. In most scenarios, this bias was more severe for the linear effects than for the interaction effect. The RMSE for conditions with misspecified scaling indicators was smallest for 2SMM, especially for low reliability scenarios, but the overall magnitude of bias was such that we cannot recommend any of the estimators in this situation. Our article showed the damage done when researchers omit cross-loadings of the scaling indicator and the importance of giving more attention to these indicators particularly if the indicators' reliability is low. It also showed that no one estimator is superior to the others across all conditions. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

Model Implied Instrumental Variables (MIIVs): An Alternative Orientation to Structural Equation Modeling.

Few dispute that our models are approximations to reality. Yet when it comes to structural equation models (SEMs), we use estimators that assume true models (e.g. maximum likelihood) and that can create biased estimates when the model is inexact. This article presents an overview of the Model Implied Instrumental Variable (MIIV) approach to SEMs from Bollen (1996). The MIIV estimator using Two Stage Least Squares (2SLS), MIIV-2SLS, has greater robustness to structural misspecifications than system wide estimators. In addition, the MIIV-2SLS estimator is asymptotically distribution free. Furthermore, MIIV-2SLS has equation-based overidentification tests that can help pinpoint misspecifications. Beyond these features, the MIIV approach has other desirable qualities. MIIV methods apply to higher order factor analyses, categorical measures, growth curve models, dynamic factor analysis, and nonlinear latent variables. Finally, MIIV-2SLS permits researchers to estimate and test only the latent variable model or any other subset of equations. In addition, other MIIV estimators beyond 2SLS are available. Despite these promising features, research is needed to better understand its performance under a variety of conditions that represent empirical applications. Empirical and simulation examples in the article illustrate the MIIV orientation to SEMs and highlight an R package MIIVsem that implements MIIV-2SLS.

Practical application of the vanishing tetrad test for causal indicator measurement models: an example from health-related quality of life.

Researchers are often faced with the task of trying to measure abstract concepts. The most common approach is to use multiple indicators that reflect an underlying latent variable. However, this 'effect indicator' measurement model is not always appropriate; sometimes the indicators instead cause the construct of interest. While the notion of 'causal indicators' has been known for some time, it is still too often ignored. However, there are limited means to determine whether a possible indicator should be treated as a cause or an effect of the latent construct of interest. Perhaps the best empirical way is to use the vanishing tetrad test (VTT), yet this method is still often overlooked. We speculate that one reason for this is the lack of published examples of its use in practice, written for an audience without extensive statistical training. The goal of this paper was to help fill this gap in the literature-to provide a basic example of how to use the VTT. We illustrated the VTT by looking at multiple items from a health related quality of life instrument that seem more likely to cause the latent variable rather than the other way around.

Socio-economic status, permanent income, and fertility: a latent-variable approach.

This paper examines how permanent income and other components of socio-economic status (SES) are related to fertility in less developed countries. Because permanent income cannot be measured directly, we employ a latent-variable method. We compare our results with those of the more common proxy-variable method and investigate the consequences of not accounting for measurement error. Using data from Ghana and Peru, we find that permanent income has a large, negative influence on fertility and that research must take the latent nature of permanent income into account to uncover its influence. Controlling for measurement error in the proxies for permanent income can also lead to substantial changes in the estimated effects of control variables. Finally, we examine which of the common proxies for permanent income most closely capture the concept. The results have implications beyond this specific dependent variable, providing evidence on the sensitivity of microanalyses to the treatment of long-term economic status.

Economic status proxies in studies of fertility in developing countries: Does the measure matter?

This paper investigates the consequences of using different economic status proxies on the estimated impact of economic status and other determinants of fertility. Using micro survey data from Ghana and Peru, we find that the proxies for income that best predict fertility are a principal components score of the ownership of consumer durable goods and a simple sum of ownership of these durable goods. Furthermore, the choice of the proxy generally has a minor influence on the predicted effects of the control variables. We compare the results from using a restricted set of proxies, such as those available in the Demographic and Health Surveys, with the results obtained using a lengthier set of proxies. Our results suggest implications beyond fertility analyses by providing researchers with an awareness of the sensitivity of microanalyses to the treatment of economic status. Our results also suggest practical recommendations for the collection of survey data.

The Noncentral Chi-square Distribution in Misspecified Structural Equation Models: Finite Sample Results from a Monte Carlo Simulation.

The noncentral chi-square distribution plays a key role in structural equation modeling (SEM). The likelihood ratio test statistic that accompanies virtually all SEMs asymptotically follows a noncentral chi-square under certain assumptions relating to misspecification and multivariate distribution. Many scholars use the noncentral chi-square distribution in the construction of fit indices, such as Steiger and Lind's (1980) Root Mean Square Error of Approximation (RMSEA) or the family of baseline fit indices (e.g., RNI, CFI), and for the computation of statistical power for model hypothesis testing. Despite this wide use, surprisingly little is known about the extent to which the test statistic follows a noncentral chi-square in applied research. Our study examines several hypotheses about the suitability of the noncentral chi-square distribution for the usual SEM test statistic under conditions commonly encountered in practice. We designed Monte Carlo computer simulation experiments to empirically test these research hypotheses. Our experimental conditions included seven sample sizes ranging from 50 to 1000, and three distinct model types, each with five specifications ranging from a correct model to the severely misspecified uncorrelated baseline model. In general, we found that for models with small to moderate misspecification, the noncentral chi-square distribution is well approximated when the sample size is large (e.g., greater than 200), but there was evidence of bias in both mean and variance in smaller samples. A key finding was that the test statistics for the uncorrelated variable baseline model did not follow the noncentral chi-square distribution for any model type across any sample size. We discuss the implications of our findings for the SEM fit indices and power estimation procedures that are based on the noncentral chi-square distribution as well as potential directions for future research.