Search

Longitudinal cohort studies provide the opportunity to examine causal effects of complex exposures on long-term health outcomes. Utilizing data from multiple cohorts has the potential to add further benefit by improving precision of estimates through data pooling and allowing examination of effect heterogeneity across contexts. However, the interpretation of findings can be complicated by biases that may be compounded when pooling data or may contribute to discrepant findings when analyses are replicated across cohorts. Here we extend the target trial framework, already well established as a powerful tool for causal inference in single-cohort studies, to address the specific challenges that can arise in the multi-cohort setting. The approach considers the target trial as a central point of reference, as opposed to comparing one study to another. This enables clear definition of the target estimand and systematic consideration of sources of bias within each cohort and additional sources of bias arising from data pooling. Consequently, analyses can be designed to reduce these biases and the resulting findings appropriately interpreted. We use a case study to demonstrate the approach and its potential to strengthen causal inference in multi-cohort studies through improved analysis design and clarity in the interpretation of findings., Comment: 34 pages, 3 figures

The aim of this research program is to analyse existing data to understand the impact of COVID-19 on the mental health and wellbeing of children, young people, and families, and how pre-pandemic experiences contribute to risk and resilience. This research program applies statistical methods to existing datasets, including cohorts collaborating through the Melbourne Children's LifeCourse Initiative. This ethics protocol provides an overview of the research program and approach to addressing ethical concerns such as secure handling of data.

The aim of the LifeCourse initiative is to enable researchers to capitalise on the availability of extensive cohort data to advance understanding of health issues emerging over the life span. To this end, LifeCourse brings together information regarding over 30 cohorts into a common platform. Each cohort is independently managed by the relevant academic or research organisation leading that study, with their own ethics and governance protocols in place. Murdoch Children’s Research Institute (MCRI) operates LifeCourse and collaborates with these academic and research organisations and their respective cohort teams to promote reuse of these valuable data, including by: presenting cohort metadata in a standardised format on a publicly accessible website, and providing a central gateway for data access requests, which are evaluated and actioned by each relevant cohort data custodian. The purpose of this document is to provide a framework for considering current data governance arrangements at the intersection between cohorts and LifeCourse, particularly regarding study metadata and data access. The primary audience are cohort data custodians and teams who undertake the day-to-day operations of their study. Articulating this framework is timely in the context of efforts across MCRI to support best practice approaches to data custodianship. We anticipate this framework will continue to evolve over time in response to changing cohort needs, Institute conventions, available technology, and best practice standards.

The LifeCourse website provides a bird’s eye view of what data have been collected across our cohorts. Measures used within a study wave are described using a set of terms to capture the constructs assessed, which are then broadly summarised into a smaller number of domains. Standardising how we describe and organise data captured across LifeCourse cohorts will allow us to achieve a more streamlined, consistent, and intuitive format for browsing and searching of cohort data elements. LifeCourse made a first significant step towards this by standardising the terms used to describe and organise the data collected across the LifeCourse cohorts. The next step required is to allocate these standardised terms to measures themselves. Rather than assigning these anew each time a measure is encountered, we aim to develop a terms-to-measures map with pre-defined term/s assigned for each measure, which will then auto populate to the website. Our previous approach when assigning LifeCourse terms to measures was ad hoc and often inconsistent. The process outlined in this report details a more methodical approach to assigning measures to the LifeCourse standard terminology, which is replicable and will lead to more consistent descriptions of the measures used across the LifeCourse cohorts.

The LifeCourse website provides a bird’s eye view of what data has been collected across our cohorts. Measures used within each study wave are described by a common set of terms that capture the constructs assessed, which are then broadly summarised into a smaller number of domains. This enables easy and intuitive browsing and searching. Historically, LifeCourse has used an in-house developed set of terms and domain groupings to describe measures. This approach was flexible and allowed data custodians to describe data in a way that they preferred. However, with the scale that the platform has grown to, this approach is no longer tenable. Of particular concern is a lack of consistency, with the same measures described using different terms between studies. Standardising the way we describe and organise data captured across LifeCourse cohorts will allow us to achieve a more streamlined, consistent, and intuitive format for browsing and searching of cohort data elements. A vital first step is to standardise the LifeCourse terminology itself – the set of terms and groupings of these terms that are used to describe and organise the data. This is the focus of this report.

The purpose of this document is to establish a data management framework for the Changing Children���s Chances (CCC) project (2021-2024) ���Child health and developmental inequities: Evidence for precision policy.��� The document aims to promote: ��� A standardised approach to data management across team members; ��� Open and transparent practices that allow other researchers to follow and replicate the work undertaken; ��� Awareness of responsibilities and requirements regarding data security and safety, with easy access to relevant information for CCC project staff; and ��� Sharing of data management approaches between researchers, capacity building and collective knowledge for the team and the wider Centre for Community Child Health.

ObjectiveThis study capitalizes on developmental data from an Australian population-based birth cohort to identify developmental markers of abnormal eating attitudes and behaviors in adolescence. The aims were twofold: (1) to develop a comprehensive path model identifying infant and childhood developmental correlates of Abnormal Eating Attitudes and Behaviors in adolescence, and (2) to explore potential gender differences.MethodData were drawn from a 30-year longitudinal study that has followed the health and development of a population based cohort across 15 waves of data collection from infancy since 1983: The Australian Temperament Project. Participants in this analysis were the 1,300 youth who completed the 11th survey at 15-16 years (1998) and who completed the eating disorder inventory at this time point.ResultsDevelopmental correlates of Abnormal Eating Attitudes and Behaviors in mid-adolescence were temperamental persistence, early gestational age, persistent high weight, teen depression, stronger peer relationships, maternal dieting behavior, and pubertal timing. Overall, these factors accounted for 28% of the variance in Abnormal Eating Attitudes and Behaviors at 15-16 years of age. Depressive symptoms, maternal dieting behavior, and early puberty were more important factors for girls. Late puberty was a more important factor for boys.DiscussionFindings address an important gap in our understanding of the etiology of Abnormal Eating Attitudes and Behaviors in adolescence and suggest multiple targets for preventive intervention.

Refereed/Peer-reviewed Background: Across the life course, socioeconomic disadvantage disproportionately afflicts those with genetic predispositions to inflammatory diseases. We describe how socioeconomic disadvantage and polygenic risk for high BMI magnify the risk of obesity across childhood, and using causal analyses, explore the hypothetical impact of intervening on socioeconomic disadvantage to reduce adolescent obesity. Methods: Data were drawn from a nationally representative Australian birth cohort, with biennial data collection between 2004 and 2018 (research and ethics committee approved). We generated a polygenic risk score for BMI using published genome-wide association studies. We measured early-childhood disadvantage (age 2–3 years) with a neighbourhood census-based measure and a family-level composite of parent income, occupation, and education. We used generalised linear regression (Poisson-log link) to estimate the risk of overweight or obesity (BMI ≥85th percentile) at age 14–15 years for children with early-childhood disadvantage (quintiles 4–5) versus average (quintile 3) and least disadvantage (quintiles 1–2), for those with high and low polygenic risk separately. Findings: For 1607 children (n=796 female, n=811 male; 31% of the original cohort [N=5107]), polygenic risk and disadvantage were both associated with overweight or obesity; effects of disadvantage were more marked as polygenic risk increased. Of children with polygenic risk higher than the median (n=805), 37% of children living in disadvantage at age 2–3 years had an overweight or obese BMI by adolescence, compared with 26% of those with least disadvantage. For genetically vulnerable children, causal analyses indicated that early neighbourhood intervention to lessen disadvantage (to quintile 1–2) would reduce risk of adolescent overweight or obesity by 23% (risk ratio 0·77; 95% CI 0·57–1·04); estimates for improving family environments were similar (0·59; 0·43–0·80). Interpretation: Actions addressing socioeconomic disadvantage could mitigate polygenic risk for developing obesity. This study benefits from population-representative longitudinal data but is limited by sample size.