Prever a suscetibilidade à dor musculoesquelética crónica em adultos jovens: olhar para trás para construir a saúde do futuro

With this project we will address current major challenges in population research on chronic pain which can only be investigated using large population-based longitudinal studies. Those challenges involve defining and refining adverse musculoskeletal pain trajectories based on data from youth and their parents, estimating the impact of childhood adversity and resilience on the embodiment of chronic musculoskeletal pain, and applying technological solutions that use emerging quantitative techniques to address classic epidemiological questions as well as the key issue of ensuring adequate data protection for research participants. Our interdisciplinary research team, together with external consultants, will examine reported and experimental data from Generation XXI, a well-established Portuguese population-based cohort of youth born in 2005/6 and followed regularly since birth, to achieve the objectives detailed below. 1. To improve the identification of children who are most likely to develop high-impact, chronic musculoskeletal pain as young adults, quantifying potential reversibility. Specifically, we will use data from four Generation XXI assessment waves from ages 7 to 18 years to: a) Identify 10-year trajectories of musculoskeletal pain experience based on questionnaire data, including estimating the reversibility of adverse pain profiles; b) Enhance differentiation among pain trajectories by adding individual and, importantly, maternal responses to two distinct but complementary modalities of quantitative sensory testing measured at ages 13 and/or 18 years, namely cuff pressure algometry and the cold pressor test; c) Build and refine statistical models to estimate the ability of the trajectories defined above to predict quality of life at age 18 years. We will have the unique opportunity to externally-validate our statistical models using data from the related Norwegian birth cohort “Tromsø Study: Fit Futures,” led by our project’s international consultant. 2. To address the underexplored role of social context in the aetiology of musculoskeletal pain trajectories. In particular, we aim to assess the impact of preventable childhood adversity and resilience on chronic musculoskeletal pain by: a) Quantifying the interaction between adverse and positive childhood experiences as a determinant of reported musculoskeletal pain trajectories, using directed acyclic graphs to depict causal assumptions, hypotheses and sources of bias; b) Testing whether there is a long-term embodiment of adversity in terms of pain sensitivity and pro-nociception from ages 13 to 18. 3. To leverage state-of-the-art machine learning techniques as novel conceptual and operational approaches to improve the conduct of population-based epidemiologic research. Specifically, we will examine the potential value of machine learning techniques as: a) Analytical tools to complement classical statistical methods, by applying unsupervised learning to address objectives 1a and 1b and supervised learning for objectives 1c, 2a, and 2b; b) Privacy-preserving approaches to the remote assessment of pain experiences, in interval cohort studies subject to attrition, using federated machine learning. Addressing these objectives will contribute to inform clinical risk stratification, public health promotion policy, and incorporation of data science techniques as part of the epidemiological toolkit.