Anna Freni-Sterrantino

In this work, we present a new approach for constructing models for covariance matrices by considering the decomposition into marginal variances and a correlation matrix. The correlation structure is deduced from a user-defined graphical structure. The graphical structure makes correlation matrices interpretable and avoids the quadratic increase of parameters as a function of the dimension. We propose an automatic approach to define a prior using a natural sequence of simpler models within the Penalized Complexity framework for the unknown parameters in these models. We illustrate this approach with simulation studies of multivariate longitudinal joint modelling, where we demonstrate some properties of the method and two real data applications: a multivariate linear regression of four biomarkers and a multivariate disease mapping. Each application underscores our method’s intuitive appeal, signifying a substantial advancement toward a more cohesive and enlightening model that facilitates a meaningful interpretation of correlation matrices.

Implement some models for correlation/covariance matrices including two approaches to model correlation matrices from a graphical structure. One use latent parent variables as proposed in Sterrantino et. al. (2024) . The other uses a graph to specify conditional relations between the variables. The graphical structure makes correlation matrices interpretable and avoids the quadratic increase of parameters as a function of the dimension. In the first approach a natural sequence of simpler models along with a complexity penalization is used. The second penalizes deviations from a base model. These can be used as prior for model parameters, considering C code through the 'cgeneric' interface for the 'INLA' package (). This allows one to use these models as building blocks combined and to other latent Gaussian models in order to build complex data models.

Forest restoration can potentially contribute to multiple global sustainable development goals. Yet, little is known about the factors associated with local actors' choice to engage in restoration, limiting the development of effective scaling strategies. Our study examines the spatial socio‐ecological factors associated with landowners' engagement in forest restoration, documented by the Atlantic Forest Restoration Pact in Brazil. We draw on Diffusion of Innovations theory to model associations between forest restoration and explanatory variables among 222,000 private properties in the Atlantic Forest. Properties with the highest cattle densities were 163.9% more likely to be restored (95% CI: 131.1%–201.3%) than those with the lowest. Large properties had a 120.0% higher restoration probability (95% CI: 90.9%–153.5%) than medium ones. Compared to reference levels, associations between forest cover (in 2010) and restoration probability were ambiguous, but properties with greater forest loss (1990–2010) were 9.0% less likely to be restored (95% CI: −12.5% to −5.3%). Properties with water bodies were 22.2% more likely to be restored (95% CI: 8.9%–37.1%), while upland properties were 46.6% less likely (95% CI: 32.3%–57.8%), and those with the highest ruggedness 22.2% less likely (95% CI: 12.4%–30.9%). Longer travel times to urban areas reduced restoration likelihood by 48.3% (95% CI: 39.7%–55.8%). Properties in the highest road distance quantile were 11.1% more likely to be restored (95% CI: 0.1%–23.2%). High forest management levels reduced restoration likelihood by 52.6% (95% CI: 15.5%–73.6%), while densely populated areas increased it by 53.8% (95% CI: 35.6%–74.4%). Properties in wealthier municipalities were more likely to be restored (54.7%, 95% CI 10.8%–116.4%). Landowners' decision‐making appears responsive to legislative requirements (on minimum forest cover and afforestation around waterbodies), underscoring their value for promoting restoration. Commercial landowners might have greater incentives to restore or be selectively targeted by restoration organisations, risking the marginalisation of smallholders in poorer areas from restoration agendas. Engagement with forest restoration is highest where there are more people and lower travel times to cities, suggesting restoration can potentially deliver ecological benefits in some of Brazil's most degraded landscapes. Read the free Plain Language Summary for this article on the Journal blog.

Composite indicators are useful for summarizing and comparing changes among different communities. The UK Office for National Statistics has created an annual England Health Index (2015–2018) comprised of three main health domains—lives, places, and people—to monitor health over time and across different geographical areas and evaluate the nation’s health. We reviewed the conceptual coherence and statistical requirements, focusing on three main steps: correlation analysis at different levels, comparison of the implemented weights, and a sensitivity and uncertainty analysis. Based on the results, we have highlighted features that have improved the statistical requirements of the forthcoming UK Health Index.

Occupation-related stress and work characteristics are possible determinants of social inequalities in epigenetic aging but have been little investigated. Here, we investigate the association of several work characteristics with epigenetic age acceleration (AA) biomarkers. The study population included employed and unemployed men and women (n = 631) from the UK Understanding Society study. We evaluated the association of employment and work characteristics related to job type, job stability; job schedule; autonomy and influence at work; occupational physical activity; and feelings regarding the job with four epigenetic age acceleration biomarkers (Hannum, Horvath, PhenoAge, GrimAge) and pace of aging (DunedinPoAm, DunedinPACE). We fitted linear regression models, unadjusted and adjusted for established risk factors, and found the following associations for unemployment (years of acceleration): HorvathAA (1.51, 95% CI 0.08, 2.95), GrimAgeAA (1.53, 95% CI 0.16, 2.90) and 3.21 years for PhenoAA (95% CI 0.89, 5.33). Job insecurity increased PhenoAA (1.83, 95% CI 0.003, 3.67), while working at night was associated with an increase of 2.12 years in GrimAgeAA (95% CI 0.69, 3.55). We found effects of unemployment to be stronger in men and effects of night shift work to be stronger in women. These results provide evidence of associations between unemployment with accelerated ageing and suggest that insecure employment and night work may also increase age acceleration. Our findings have implications for policies relating to current changes in working conditions and highlight the utility of biological age biomarkers in studies in younger populations without long-term health information.

Recent evidence indicates consistent association of low socioeconomic status with epigenetic age acceleration, measured from DNA methylation. As work characteristics and job stressors are crucial components of socioeconomic status, we investigated their association with various measures of epigenetic age acceleration. The study population included employed and unemployed men and women (n=604) from the Northern Finland Birth Cohort 1966. We investigated the association of job strain, effort-reward imbalance and work characteristics with five biomarkers of epigenetic aging (Hannum, Horvath, PhenoAge, GrimAge, and DunedinPoAm). Our results indicate few significant associations between work stress indicators and epigenetic age acceleration, limited to a range of ±2 years, and smoking recording the highest effect on GrimAge age acceleration biomarker between current and no smokers (median difference 4.73 years (IQR 1.18, 8.41). PhenoAgeAA was associated with job strain active work (β=-1.301 95%CI -2.391, -0.212), slowing aging of less than 1.5 years, and working as white-collar slowed aging six months (GrimAgeAA β=-0.683, 95%CI -1.264, -0.102) when compared to blue collars. Association was found for working for more than 40 hours per week that increased the aging over 1.5 years, (HorvathAA β =2.058 95%CI 0.517,3.599, HannumAA β=1.567, 95%CI 0.415,2.719). The pattern of associations was different between women and men and some of the estimated effects are inconsistent with current literature. Our results provide the first evidence of association of work conditions with epigenetic aging biomarkers. However, further epidemiological research is needed to fully understand how work-related stress affects epigenetic age acceleration in men and women in different societies.

A Plea for the Need to Investigate the Health Effects of Gig-Economy

Objectives: To investigate long-term associations between metal components of particulate matter (PM) and mortality and lung cancer incidence. Design: Small area (ecological) study. Setting: Population living in all wards (~9000 individuals per ward) in the London and Oxford area of England, comprising 13.6 million individuals. Exposure and outcome measures: We used land use regression models originally used in the Transport related Air Pollution and Health Impacts—Integrated Methodologies for Assessing Particulate Matter study to estimate exposure to copper, iron and zinc in ambient air PM. We examined associations of metal exposure with Office for National Statistics mortality data from cardiovascular disease (CVD) and respiratory causes and with lung cancer incidence during 2008–2011. Results: There were 108 478 CVD deaths, 48 483 respiratory deaths and 24 849 incident cases of lung cancer in the study period and area. Using Poisson regression models adjusted for area-level deprivation, tobacco sales and ethnicity, we found associations between cardiovascular mortality and PM _2.5 copper with interdecile range (IDR 2.6–5.7 ng/m ³ ) and IDR relative risk (RR) 1.005 (95%CI 1.001 to 1.009) and between respiratory mortality and PM ₁₀ zinc (IDR 1135–153 ng/m ³ ) and IDR RR 1.136 (95%CI 1.010 to 1.277). We did not find relevant associations for lung cancer incidence. Metal elements were highly correlated. Conclusion: Our analysis showed small but not fully consistent adverse associations between mortality and particulate metal exposures likely derived from non-tailpipe road traffic emissions (brake and tyre wear), which have previously been associated with increases in inflammatory markers in the blood.