Longitudinal Studies

ObjectivesMotivated by recent calls to use electronic health records for research, we reviewed the application and development of methods for addressing the bias from unmeasured confounding in longitudinal data.Study design and settingMethodological review of existing literature. We searched MEDLINE and EMBASE for articles addressing the threat to causal inference from unmeasured confounding in nonrandomized longitudinal health data through quasi-experimental analysis.ResultsAmong the 121 studies included for review, 84 used instrumental variable analysis (IVA), of which 36 used lagged or historical instruments. Difference-in-differences (DiD) and fixed effects (FE) models were found in 29 studies. Five of these combined IVA with DiD or FE to try to mitigate for time-dependent confounding. Other less frequently used methods included prior event rate ratio adjustment, regression discontinuity nested within pre-post studies, propensity score calibration, perturbation analysis, and negative control outcomes.ConclusionWell-established econometric methods such as DiD and IVA are commonly used to address unmeasured confounding in nonrandomized longitudinal studies, but researchers often fail to take full advantage of available longitudinal information. A range of promising new methods have been developed, but further studies are needed to understand their relative performance in different contexts before they can be recommended for widespread use.

BackgroundThe extent to which omega-3 fatty acid status is related to risk for death from any cause and for incident cardiovascular disease (CVD) remains controversial.ObjectiveTo examine these associations in the Framingham Heart Study.DesignProspective and observational.SettingFramingham Heart Study Offspring cohort.MeasurementsThe exposure marker was red blood cell levels of eicosapentaenoic and docosahexaenoic acids (the Omega-3 Index) measured at baseline. Outcomes included mortality (total, CVD, cancer, and other) and total CVD events in participants free of CVD at baseline. Follow-up was for a median of 7.3 years. Cox proportional hazards models were adjusted for 18 variables (demographic, clinical status, therapeutic, and CVD risk factors).ResultsAmong the 2500 participants (mean age 66 years, 54% women), there were 350 deaths (58 from CVD, 146 from cancer, 128 from other known causes, and 18 from unknown causes). There were 245 CVD events. In multivariable-adjusted analyses, a higher Omega-3 Index was associated with significantly lower risks (P-values for trends across quintiles) for total mortality (P = .02), for non-CVD and non-cancer mortality (P = .009), and for total CVD events (P = .008). Those in the highest (>6.8%) compared to those in the lowest Omega-3 Index quintiles (LimitationsRelatively short follow-up time and one-time exposure assessment.ConclusionsA higher Omega-3 Index was associated with reduced risk of both CVD and all-cause mortality.

Over the life course, we are invariably faced with some form of adversity. The process of positively adapting to adverse events is known as 'resilience'. Despite the acknowledgement of 2 common components of resilience, that is, adversity and positive adaptation, no consensus operational definition has been agreed. Resilience operationalisations have been reviewed in a cross-sectional context; however, a review of longitudinal methods of operationalising resilience has not been conducted. The present study conducts a systematic review across Scopus and Web of Science capturing studies of ageing that posited operational definitions of resilience in longitudinal studies of ageing. Thirty-six studies met inclusion criteria. Non-acute events, for example, cancer, were the most common form of adversity identified and psychological components, for example, the absence of depression, the most common forms of positive adaptation. Of the included studies, 4 used psychometrically driven methods, that is, repeated administration of established resilience metrics, 9 used definition-driven methods, that is, a priori establishment of resilience components and criteria, and 23 used data-driven methods, that is, techniques that identify resilient individuals using latent variable models. Acknowledging the strengths and limitations of each operationalisation is integral to the appropriate application of these methods to life course and longitudinal resilience research.

ObjectiveAs more researchers employ longitudinal research designs, which integrate multiple methods and multiple (fixed) situations, the need for appropriate analytical methods arises.MethodMultimethod latent state-trait models for random and fixed situations (MM-LST-RF; Hintz et al. 2019) provide a means with which person characteristics, (fixed) situation, and method effects, as well as their interactions can be studied. While these models are very versatile, their complexity poses a significant hurdle to their implementation.ResultsThis tutorial helps facilitate the application of MM-LST-RF models. First, we present two simpler methodological approaches in which the full MM-LST-RF model is broken down into its (a) multimethod and (b) random and fixed situation components. Key parameters and model coefficients are highlighted using a motivational example. Second, we present a user-friendly shiny app based on a newly developed R function. Users are walked through the process of specifying, estimating, and interpreting an MM-LST-RF model guided by detailed explanations of all specification options and practical use recommendations.ConclusionThe shiny app facilitates the analysis of data from longitudinal study designs implementing multiple methods and (fixed) situations, helping researchers gain a deeper understanding of psychological constructs.