Explore the vast range of titles available.

Background Collection of intensive longitudinal health outcomes allows joint modeling of their mean (location) and variability (scale). Focusing on the location of the outcome, measures to detect influential subjects in longitudinal data using standard mixed-effects regression models (MRMs) have been widely discussed. However, no existing approach enables the detection of subjects that heavily influence the scale of the outcome. Methods We propose applying mixed-effects location scale (MELS) modeling combined with commonly used influence measures such as Cook’s distance and DFBETAS to fill this gap. In this paper, we provide a framework for researchers to follow when trying to detect influential subjects for both the scale and location of the outcome. The framework allows detailed examination of each subject’s influence on model fit as well as point estimates and precision of coefficients in different components of a MELS model. Results We simulated two common scenarios in longitudinal healthcare studies and found that influence measures in our framework successfully capture influential subjects over 99% of the time. We also re-analyzed data from a health behavior study and found 4 particularly influential subjects, among which two cannot be detected by influence analyses via regular MRMs. Conclusion The proposed framework can help researchers detect influential subject(s) that will be otherwise overlooked by influential analysis using regular MRMs and analyze all data in one model despite influential subjects.

Poor diet, low physical activity, high stress, and poor sleep are prevalent modifiable risk factors for chronic diseases. Synergistic effects of interventions targeting diet and activity on other health risk behaviors such as stress and sleep are understudied. To conduct a secondary data analysis to investigate whether interventions targeting diet and activity produce collateral improvements in stress and sleep. Make Better Choices 2 was a randomized clinical trial to test a technology-assisted coaching intervention with modest incentives to improve diet and activity, as compared to a matched intervention targeting improved stress and sleep. Participants (n = 212) were adults (76.4% female, 59% non-white minority, mean age = 40.8 years) with multiple diet and activity risk behaviors. For 7 days at baseline, 3-, 6-, and 9-months, participants reported perceived sleep duration, stress, diet, and activity through a smartphone application. Outcomes were evaluated by linear mixed models. The study was registered on clinicaltrials.gov (NCT01249989). Both interventions produced significant, statistically comparable improvements in average daily stress rating (z = 1.35, p = .177). Reduction in average daily stress rating was 1.68 following diet/activity intervention (z = -12.25, P < .001) and 2.08 following stress/sleep intervention (z = -7.83, P < .001) on an 11-point Likert scale. Though changes in sleep duration for both groups were clinically meaningful, the stress/sleep intervention produced statistically larger improvements in sleep as compared to the diet/activity intervention (z = -3.79, P < .001). Sleep duration increased 26.39 minutes following diet/activity intervention (z = 3.16, P = 0.002) and 92.65 minutes following stress/sleep intervention (z = 5.912, P < 0.001). Findings suggest that diet and activity behavior change interventions can effectively improve outcomes within and between the behavior domains they directly target. Future research should identify common mechanistic pathways to inform development of interventions that efficiently change multiple health behaviors implicated in chronic disease morbidity and mortality.

Purpose To evaluate weight change patterns over time following the diagnosis of breast cancer and to examine the association of post-diagnosis weight change and survival outcomes in Black and White patients. Methods The study included 2888 women diagnosed with non-metastatic breast cancer in 2000–2017 in Chicago. Longitudinal repeated measures of weight and height were collected, along with a questionnaire survey including questions on body size. Multilevel mixed-effects models were used to examine changes in body mass index (BMI). Delayed entry Cox proportional hazards models were used to investigate the impacts of changing slope of BMI on survival outcomes. Results At diagnosis, most patients were overweight or obese with a mean BMI of 27.5 kg/m 2 and 31.5 kg/m 2 for Blacks and Whites, respectively. Notably, about 45% of the patients had cachexia before death and substantial weight loss started about 30 months before death. In multivariable-adjusted analyses, compared to stable weight, BMI loss (> 0.5 kg/m 2 /year) showed greater than 2-fold increased risk in overall survival (hazard ratio [HR] = 2.60, 95% CI 1.88–3.59), breast cancer-specific survival (HR = 3.05, 95% CI 1.91–4.86), and disease-free survival (HR = 2.12, 95% CI 1.52–2.96). The associations were not modified by race, age at diagnosis, and pre-diagnostic weight. BMI gain (> 0.5 kg/m 2 /year) was also related to worse survival, but the effect was weak (HR = 1.60, 95% CI 1.10–2.33 for overall survival). Conclusion BMI loss is a strong predictor of worse breast cancer outcomes. Growing prevalence of obesity may hide diagnosis of cancer cachexia, which can occur in a large proportion of breast cancer patients long before death.

Prevalent co-occurring poor diet and physical inactivity convey chronic disease risk to the population. Large magnitude behavior change can improve behaviors to recommended levels, but multiple behavior change interventions produce small, poorly maintained effects. The Make Better Choices 2 trial tested whether a multicomponent intervention integrating mHealth, modest incentives, and remote coaching could sustainably improve diet and activity. Between 2012 and 2014, the 9-month randomized controlled trial enrolled 212 Chicago area adults with low fruit and vegetable and high saturated fat intakes, low moderate to vigorous physical activity (MVPA) and high sedentary leisure screen time. Participants were recruited by advertisements to an open-access website, screened, and randomly assigned to either of two active interventions targeting MVPA simultaneously with, or sequentially after other diet and activity targets (N=84 per intervention) or a stress and sleep contact control intervention (N=44). They used a smartphone app and accelerometer to track targeted behaviors and received personalized remote coaching from trained paraprofessionals. Perfect behavioral adherence was rewarded with an incentive of US $5 per week for 12 weeks. Diet and activity behaviors were measured at baseline, 3, 6, and 9 months; primary outcome was 9-month diet and activity composite improvement. Both simultaneous and sequential interventions produced large, sustained improvements exceeding control (P<.001), and brought all diet and activity behaviors to guideline levels. At 9 months, the interventions increased fruits and vegetables by 6.5 servings per day (95% CI 6.1-6.8), increased MVPA by 24.7 minutes per day (95% CI 20.0-29.5), decreased sedentary leisure by 170.5 minutes per day (95% CI -183.5 to -157.5), and decreased saturated fat intake by 3.6% (95% CI -4.1 to -3.1). Retention through 9-month follow-up was 82.1%. Self-monitoring decreased from 96.3% of days at baseline to 72.3% at 3 months, 63.5% at 6 months, and 54.6% at 9 months (P<.001). Neither attrition nor decline in self-monitoring differed across intervention groups. Multicomponent mHealth diet and activity intervention involving connected coaching and modest initial performance incentives holds potential to reduce chronic disease risk. ClinicalTrials.gov NCT01249989; https://clinicaltrials.gov/ct2/show/NCT01249989 (Archived by WebCite at https://clinicaltrials.gov/ct2/show/NCT01249989).

For longitudinal data, mixed models include random subject effects to indicate how subjects influence their responses over repeated assessments. The error variance and the variance of the random effects are usually considered to be homogeneous. These variance terms characterize the within-subjects (i.e., error variance) and between-subjects (i.e., random-effects variance) variation in the data. In studies using ecological momentary assessment (EMA), up to 30 or 40 observations are often obtained for each subject, and interest frequently centers around changes in the variances, both within and between subjects. In this article, we focus on an adolescent smoking study using EMA where interest is on characterizing changes in mood variation. We describe how covariates can influence the mood variances, and also extend the standard mixed model by adding a subject-level random effect to the within-subject variance specification. This permits subjects to have influence on the mean, or location, and variability, or (square of the) scale, of their mood responses. Additionally, we allow the location and scale random effects to be correlated. These mixed-effects location scale models have useful applications in many research areas where interest centers on the joint modeling of the mean and variance structure.

Background Longitudinal assessments of usage are often conducted for multiple substances (e.g., cigarettes, alcohol and marijuana) and research interests are often focused on the inter-substance association. We propose a multivariate longitudinal modeling approach for jointly analyzing the ordinal multivariate substance use data. Methods We describe how the binary random slope logistic regression model can be extended to the multi-category ordinal outcomes. We also describe how the proportional odds assumption can be relaxed by allowing differential covariate effects on different cumulative logits for multiple outcomes. Data are analyzed from a P01 study that evaluates the usage levels of cigarettes, alcohol and marijuana repeatedly across 8 measurement waves during 7 consecutive years. Results 1263 subjects participated in the study with informed consent, among whom 56.6% are females. Males and females show significant differences in terms of the time trend for substance use. Specifically, males showed steeper trends on cigarette and marijuana use over time compared to females, while less so for alcohol. For all three substances, age effects appear to be different for different cumulative logits, indicating the violation of proportional odds assumption. Conclusions The multivariate mixed cumulative logit model offers the most flexibility and allows one to examine the inter-substance association when proportional odds assumption is violated.

Walking cadence has been suggested as a measure of activity intensity; however, it remains uncertain if prefrail and frail older adults can increase their walking cadence and if doing so leads to improvements in functional capacity. We aimed to determine if cadence can be increased and if this leads to improvement in functional capacity in prefrail and frail older adults. We performed a secondary data analysis of a walking intervention in prefrail and frail older adults living in retirement communities. Patients were randomized to Casual Speed Walking (CSW) and High-Intensity Walking (HIW) groups. Our primary outcome was improvement in 6-minute walk test distance above the minimally clinical important difference. We performed linear and logistic mixed-effects regressions to analyze our aims. 102 participants were included in the final analysis with 56 in the CSW group and 46 in the HIW group. Participants in the HIW group increased their walking cadence as compared to the CSW group during the intervention (HIW 100[88, 111] steps/min vs. CSW 77[65, 86] steps/min; P < 0.001). Participants that increased their walking cadence demonstrated an increased odds of improvement in their 6-minute walk test minimum clinically important difference (OR: 0.11, 95% CI: 0.033, 0.18; p = 0.005). Older adults can increase their walking cadence and walking cadence can serve as a surrogate measure of activity intensity during walking interventions. An increase of 14 steps/minute from their comfortable walking cadence increased the odds of improvement in 6-minute walk test minimum clinically important difference.

Cigarette smoking is highly concentrated among individuals with lower socioeconomic status (SES) who often lack access to smoking cessation services. Thus, smoking cessation in lower SES adults remains a critical public health concern that warrants further study and attention. Smokers attempting to quit are at the highest risk for lapse within the first weeks of their quit attempt, and an initial lapse is highly likely to lead to full relapse. It is essential to identify and understand behavioral factors that may increase or decrease the likelihood of successful smoking cessation among lower SES adults during a quit attempt (pre-and post-quit). Recently, sleep dysregulation, such as insufficient sleep duration, has been considered as a potential intervention target to address smoking behaviors (e.g., number of cigarettes smoked per day) and improve smoking cessation outcomes (e.g., abstinence). Recent studies have found that lower SES is associated with higher rates of poor sleep. Thus, SES should be accounted for when assessing sleep dysregulation during smoking cessation attempts. Although previous studies have examined the relationship between sleep dysregulation and smoking behavior and/or cessation outcomes, they have several methodological limitations, including the use of retrospective survey methods, use of cross-sectional study designs, relying solely on laboratory-based data collection, not assessing integrated sleep health dimensions (usually only sleep duration or quality is assessed), omitting lower SES adults who smoke, and focusing on a single pathway rather than bidirectional associations. This study will use a real-time data capture approach among lower SES adults who are attempting to quit smoking. This approach will involve a granular examination of the bidirectional and temporal associations between daily sleep dysregulation and smoking cession processes (pre- and post-quit) using smartphone-based ecological momentary assessment (EMA) and wearable sensors. Specifically, we aim to identify bidirectional and temporal associations between daily smoking abstinence and sleep dysregulation via EMA and wrist-worn sensors during the first four weeks of a smoking cessation attempt. Findings from this study will yield preliminary data that will be used to develop and implement a Just-in-Time-Adaptive Intervention (JITAI) that aims to improve sleep health during smoking cessation.

Objectives To examine whether long-term air pollution exposure is associated with central hemodynamic and brachial artery stiffness parameters. Methods We assessed central hemodynamic parameters including central blood pressure, cardiac parameters, systemic vascular compliance and resistance, and brachial artery stiffness measures [including brachial artery distensibility (BAD), compliance (BAC), and resistance (BAR)] using waveform analysis of the arterial pressure signals obtained from a standard cuff sphygmomanometer (DynaPulse2000A, San Diego, CA). The long-term exposures to particles with an aerodynamic diameter < 2.5 μm (PM2.5) and nitrogen dioxide (NO2) for the 3-year periods prior to enrollment were estimated at residential addresses using fine-scale intra-urban spatiotemporal models. Linear mixed models adjusted for potential confounders were used to examine associations between air pollution exposures and health outcomes. Results The cross-sectional study included 2,387 Chicago residents (76% African Americans) enrolled in the ChicagO Multiethnic Prevention And Surveillance Study (COMPASS) during 2013–2018 with validated address information, PM2.5 or NO2, key covariates, and hemodynamics measurements. We observed long-term concentrations of PM2.5 and NO2 to be positively associated with central systolic, pulse pressure and BAR, and negatively associated with BAD, and BAC after adjusting for relevant covariates. A 1-µg/m 3 increment in preceding 3-year exposures to PM2.5 was associated with 1.8 mmHg higher central systolic (95% CI: 0.98, 4.16), 1.0 mmHg higher central pulse pressure (95% CI: 0.42, 2.87), a 0.56%mmHg lower BAD (95% CI: -0.81, -0.30), and a 0.009 mL/mmHg lower BAC (95% CI: -0.01, -0.01). Conclusion This population-based study provides evidence that long-term exposures to PM2.5 and NO2 is related to central BP and arterial stiffness parameters, especially among African Americans.

This paper discusses statistical models for multilevel ordinal data that may be more appropriate for prevention outcomes than models that assume continuous measurement and normality. Prevention outcomes often have distributions that make them inappropriate for many popular statistical models that assume normality and are more appropriately considered ordinal outcomes. Despite this, the modeling of ordinal outcomes is often not well understood. This article discusses ways to analyze multilevel ordinal outcomes that are clustered or longitudinal, including the proportional odds regression model for ordinal outcomes, which assumes that the covariate effects are the same across the levels of the ordinal outcome. The article will cover how to test this assumption and what to do if it is violated. It will also discuss application of these models using computer software programs.