Graves' disease is routinely treated with antithyroid drugs, radioiodine, or surgery, but whether the choice of initial therapy influences long-term outcomes is uncertain. We evaluated cardiovascular morbidity and mortality according to the method and effectiveness of primary therapy in Graves' disease. In this retrospective cohort study, we identified patients with hyperthyroidism, diagnosed between Jan 1, 1998, and Dec 31, 2013, from a thyroid-stimulating hormone (TSH)-receptor antibody (TRAb) test register in south Wales, UK, and imported their clinical data into the All-Wales Secure Anonymised Information Linkage (SAIL) Databank (Swansea University, Swansea, UK). Patients with Graves' disease, defined by positive TRAb tests, were selected for the study, and their clinical data were linked with outcomes in SAIL. We had no exclusion criteria. Patients were matched by age and sex to a control population (1:4) in the SAIL database. Patients were grouped by treatment within 1 year of diagnosis into the antithyroid drug group, radioiodine with resolved hyperthyroidism group (radioiodine group A), or radioiodine with unresolved hyperthyroidism group (radioiodine group B). We used landmark Kaplan-Meier and Cox regression models to analyse the association of treatment with the primary outcome of all-cause mortality and the secondary outcome of major adverse cardiovascular events (myocardial infarction, heart failure, ischaemic stroke, or death) with the landmark set at 1 year after diagnosis. We analysed the association between outcomes and concentration of TSH using Cox regression and outcomes and free thyroxine (FT4) concentration using restricted cubic-spline regression models. We extracted patient-level data on 4189 patients (3414 [81·5%] females and 775 [18·5%] males) with Graves' disease and 16 756 controls (13 656 [81·5%] females and 3100 [18·5%] males). In landmark analyses, 3587 patients were in the antithyroid drug group, 250 were in radioiodine group A, 182 were in radioiodine group B. Patients had increased all-cause mortality compared with controls (hazard ratio [HR] 1·22, 95% CI 1·05-1·42). Compared with patients in the antithyroid drug group, mortality was lower among those in radioiodine group A (HR 0·50, 95% CI 0·29-0·85), but not for those in radioiodine group B (HR 1·51, 95% CI 0·96-2·37). Persistently low TSH concentrations at 1 year after diagnosis were associated with increased mortality independent of treatment method (HR 1·55, 95% CI 1·08-2·24). Spline regressions showed a positive non-linear relationship between FT4 concentrations at 1 year and all-cause mortality. Regardless of the method of treatment, early and effective control of hyperthyroidism among patients with Graves' disease is associated with improved survival compared with less effective control. Rapid and sustained control of hyperthyroidism should be prioritised in the management of Graves' disease and early definitive treatment with radioiodine should be offered to patients who are unlikely to achieve remission with antithyroid drugs alone. National Institute for Social Care and Health Research, Wales.

A differentially methylated region (DMR) is a genomic region in which DNA methylation is consistently positively or negatively associated with a phenotype or exposure. We demonstrate that existing algorithms for identifying DMRs either fail to consistently control false positive rates (comb-p and DMRcate), suffer from low power (bumphunter) or lack modeling flexibility (seqlm). We introduce a new method, dmrff, that overcomes these shortcomings and can additionally be used to meta-analyze multiple datasets. When applied to investigate associations of age in multiple datasets, dmrff identifies novel DMRs near genes previously linked to age. An R implementation is available on Github (http://github.com/perishky/dmrff).

The hostile environment that older lesbian, gay, bisexual and transgender (LGBT) people faced at younger ages in the United Kingdom (UK) may have a lasting negative impact on their health. This systematic scoping review adds to the current knowledge base through comprehensively synthesising evidence on what is known about the extent and nature of health and care inequalities, as well as highlighting gaps in the evidence which point the way towards future research priorities. We searched four databases, undertook manual searching, and included studies which presented empirical findings on LGBT people aged 50+ in the UK and their physical and mental health or social care status. From a total of 5,738 records, 48 papers from 42 studies were eligible and included for data extraction. The synthesis finds that inequities exist across physical and mental health, as well as in social care, exposure to violence and loneliness. Social care environments appeared as a focal point for inequities and formal care environments severely compromised the identity and relationships that older LGBT people developed over their lifecourse. Conversely, the literature demonstrated how some older LGBT people successfully negotiated age-related transitions,e.g.emphasising the important role of LGBT-focused social groups in offsetting social isolation and loneliness. While there exist clear policy implications around the requirement for formal care environments to change to accommodate an increasingly diverse older population, there is also a need to explore how to support older LGBT people to maintain their independence for longer, reducing the need for formal care.

Applications of multilevel models to continuous outcomes nearly always assume constant residual variance and constant random effects variances and covariances. However, modeling heterogeneity of variance can prove a useful indicator of model misspecification, and in some educational and behavioral studies, it may even be of direct substantive interest. The purpose of this article is to review, describe, and illustrate a set of recent extensions to two-level models that allow the residual and random effects variance–covariance components to be specified as functions of predictors. These predictors can then be entered with random coefficients to allow the Level-1 heteroscedastic relationships to vary across Level-2 units. We demonstrate by simulation that ignoring Level-2 variability in residual variances leads the Level-1 variance function regression coefficients to be estimated with spurious precision. We discuss software options for fitting these extensions, and we illustrate them by reanalyzing the classic High School and Beyond data and two-level school effects models presented by Raudenbush and Bryk.

Background: Effects of labor force participation on mental health can be difficult to discern due to the possibility of selection bias. Previous research typically adjusts for direct selection (reverse causality) but ignores indirect selection (unmeasured confounders). Methods: We investigate the relationship between men’s employment transitions and mental health using a dynamic simultaneous equations model applied to data from the British Household Panel Survey (1991–2009). Outcome is self-reported distress and anxiety as summed on a 12-point scale. We allow for direct selection by allowing prior mental health to affect both subsequent mental health and employment transitions in the joint model. We adjust for indirect selection by allowing for residual correlation between mental health and employment. Results: Moving from unemployment to employment was strongly associated with an improvement in mental health, whereas becoming unemployed was detrimental. However, these associations were attenuated by unmeasured confounders. After adjustment for indirect selection, the increased distress and anxiety associated with becoming unemployed decreased from 2.5 (95% confidence interval = 2.2 to 2.7) to 2.2 (2.0 to 2.5). (A change of 2.5 equates to half a standard deviation on the 12-point scale.) The improvement with moving from unemployment to employment was also weakened slightly (from −2.1 [−2.4 to −1.7] to −1.8 [−2.1 to −1.5]). Conclusions: There was strong evidence of indirect selection, but less support for direct selection. Nevertheless, the effects on psychological health of transitions between employment and unemployment, and between employment and economic inactivity, remained substantial after adjusting for selection.

Objectives: Modeling the health and care trajectories of lesbian, gay, and bisexual (LGB) is essential to identify inequalities and support needs, yet because of the small sample of LGB people in any one survey, current evidence relies on studies that have poor generalizability and low power. This study assesses the magnitude of health inequalities among older LGB people across 10 outcomes, informed by evidence on the health trajectories and distinct LGB history of the United Kingdom. Method: A systematic review was conducted of representative data sources on older LGB and heterosexual people’s health and care status in the United Kingdom. Individual Participant Data (IPD) meta-analysis was employed to synthesize data from up to 25 different sources. To account for the intricacies of individual data sets, the analysis employed a two-stage approach where an odds ratio and standard error was calculated for each data set individually, before being meta-analyzed through DerSimonian and Laird random effects models. Results: Among men aged 50+, being gay, bisexual, or having another nonheterosexual orientation is associated with an increased risk of reporting long-term illness and health-related limitations. Indicators of mental health also suggest that gay and bisexual men are more likely to report low life satisfaction and to have attempted suicide over their life time. Among women, differences are apparent with regards to self-rated health as well as with engagement with risky health behaviors. Discussion: The findings corroborate the minority stress theory, but they also generate new questions for researchers around when and how these inequalities emerge.