madman
Super Moderator
Androgens In Men Study (AIMS): protocol for meta-analyses of individual participant data investigating associations of androgens with health outcomes in men
Abstract
Introduction This study aims to clarify the role(s) of endogenous sex hormones to influence health outcomes in men, specifically to define the associations of plasma testosterone with the incidence of cardiovascular events, cancer, dementia, and mortality risk, and to identify factors predicting testosterone concentrations. Data will be accrued from at least three Australian, two European, and four North American population-based cohorts involving approximately 20 000 men.
Methods and analysis Eligible studies include prospective cohort studies with baseline testosterone concentrations measured using mass spectrometry and 5 years of follow-up data on incident cardiovascular events, mortality, cancer diagnoses or deaths, new-onset dementia, or decline in cognitive function recorded. Data for men, who were not taking androgens or drugs suppressing testosterone production, metabolism, or action; and had no prior orchidectomy, are eligible. Systematic literature searches were conducted from 14 June 2019 to 31 December 2019, with no date range set for searches. Aggregate level data will be sought where individual participant data (IPD) are not available. One-stage IPD random-effects meta-analyses will be performed, using linear mixed models, generalized linear mixed models, and either stratified or frailty-augmented Cox regression models. Heterogeneity in estimates from different studies will be quantified and bias investigated using funnel plots. Effect size estimates will be presented in forest plots and non-negligible heterogeneity and bias investigated using subgroup or meta-regression analyses.
Ethics and dissemination Ethics approvals obtained for each of the participating cohorts state that participants have consented to have their data collected and used for research purposes. The Androgens In Men Study has been assessed as exempt from ethics review by the Human Ethics office at the University of Western Australia (file reference number RA/4/20/5014). Each of the component studies had obtained ethics approvals; please refer to respective component studies for details. Research findings will be disseminated to the scientific and broader community via the publication of four research articles, with each involving a separate set of IPD meta-analyses (articles will investigate different, distinct outcomes), at scientific conferences and meetings of relevant professional societies. Collaborating cohort studies will disseminate findings to study participants and local communities.
Discussion
Although several published meta-analyses have investigated associations of endogenous testosterone with health outcomes in men,75–86 none have conducted IPD metaanalyses of health outcomes as planned for this study. A previous IPD meta-analysis focused on the outcome of metabolic syndrome.78 Results from this study will improve on previously published estimates from individual studies, in terms of the generalisability of findings. Estimates from the IPD meta-analyses are also likely to be more reliable than those published from conventional meta-analyses because they typically have higher statistical power and provide scope for controlling for important confounders and risk factors.37 38 Uncertainty will undoubtedly remain due to the possibility of confounding influences of unadjusted effects. However, unlike a randomized controlled trial, this study avoids the need to subject individuals to interventions and provides a more comprehensive characterization of temporal relationships between baseline testosterone concentrations and a range of key health outcomes.87
Accordingly, it is hoped that the AIMS collaboration will ultimately complement the research efforts and outputs from multiple prospective cohort studies by drawing on the collective body of evidence to clarify the role of endogenous sex hormone levels on major health outcomes in men. It is possible that this work might also elucidate new understanding, arising from the improved scope for fitting more complex models due to increased statistical power or from patterns detected in subgroup or meta-regression analyses. Clinically, research outputs will be used to identify the scope and optimal recruitment criteria for future trials of testosterone therapy. These data will also allow reference ranges for testosterone in men across ages and geographical locations to be refined, to inform recommendations for clinical practice more generally.
Abstract
Introduction This study aims to clarify the role(s) of endogenous sex hormones to influence health outcomes in men, specifically to define the associations of plasma testosterone with the incidence of cardiovascular events, cancer, dementia, and mortality risk, and to identify factors predicting testosterone concentrations. Data will be accrued from at least three Australian, two European, and four North American population-based cohorts involving approximately 20 000 men.
Methods and analysis Eligible studies include prospective cohort studies with baseline testosterone concentrations measured using mass spectrometry and 5 years of follow-up data on incident cardiovascular events, mortality, cancer diagnoses or deaths, new-onset dementia, or decline in cognitive function recorded. Data for men, who were not taking androgens or drugs suppressing testosterone production, metabolism, or action; and had no prior orchidectomy, are eligible. Systematic literature searches were conducted from 14 June 2019 to 31 December 2019, with no date range set for searches. Aggregate level data will be sought where individual participant data (IPD) are not available. One-stage IPD random-effects meta-analyses will be performed, using linear mixed models, generalized linear mixed models, and either stratified or frailty-augmented Cox regression models. Heterogeneity in estimates from different studies will be quantified and bias investigated using funnel plots. Effect size estimates will be presented in forest plots and non-negligible heterogeneity and bias investigated using subgroup or meta-regression analyses.
Ethics and dissemination Ethics approvals obtained for each of the participating cohorts state that participants have consented to have their data collected and used for research purposes. The Androgens In Men Study has been assessed as exempt from ethics review by the Human Ethics office at the University of Western Australia (file reference number RA/4/20/5014). Each of the component studies had obtained ethics approvals; please refer to respective component studies for details. Research findings will be disseminated to the scientific and broader community via the publication of four research articles, with each involving a separate set of IPD meta-analyses (articles will investigate different, distinct outcomes), at scientific conferences and meetings of relevant professional societies. Collaborating cohort studies will disseminate findings to study participants and local communities.
Discussion
Although several published meta-analyses have investigated associations of endogenous testosterone with health outcomes in men,75–86 none have conducted IPD metaanalyses of health outcomes as planned for this study. A previous IPD meta-analysis focused on the outcome of metabolic syndrome.78 Results from this study will improve on previously published estimates from individual studies, in terms of the generalisability of findings. Estimates from the IPD meta-analyses are also likely to be more reliable than those published from conventional meta-analyses because they typically have higher statistical power and provide scope for controlling for important confounders and risk factors.37 38 Uncertainty will undoubtedly remain due to the possibility of confounding influences of unadjusted effects. However, unlike a randomized controlled trial, this study avoids the need to subject individuals to interventions and provides a more comprehensive characterization of temporal relationships between baseline testosterone concentrations and a range of key health outcomes.87
Accordingly, it is hoped that the AIMS collaboration will ultimately complement the research efforts and outputs from multiple prospective cohort studies by drawing on the collective body of evidence to clarify the role of endogenous sex hormone levels on major health outcomes in men. It is possible that this work might also elucidate new understanding, arising from the improved scope for fitting more complex models due to increased statistical power or from patterns detected in subgroup or meta-regression analyses. Clinically, research outputs will be used to identify the scope and optimal recruitment criteria for future trials of testosterone therapy. These data will also allow reference ranges for testosterone in men across ages and geographical locations to be refined, to inform recommendations for clinical practice more generally.
Last edited:
