Assessing environmental and pandemic influences on mortality through seasonal time series models

Understanding the interplay between air pollution, climate variables, and mortality is essential for developing evidence-based public health policies and mitigating environmental health risks. Seasonal mortality patterns, particularly winter peaks, are often associated with increased air pollution and low temperatures. However, understanding the combined associations of these factors, especially under global disruptions such as the COVID-19 pandemic, poses a challenge for public health research and policy. This study shows that the monthly number of deaths from all-natural causes in Portugal from 2010 to 2022 exhibits a pronounced seasonal pattern, with winter mortality about 19 % higher than the annual mean. In the same period, air pollutant concentrations show clear seasonality, with NO₂ (+24 %) and PM₁₀ (+12 %) peaking in winter, while minimum temperatures fall about 42 % below the mean. Conversely, ozone exposure (SOMO35) reaches highest levels in summer, reflecting distinct seasonal burden. Using Seasonal Autoregressive Integrated Moving Average models with Predictor Variables (SARIMAX), the analysis identifies complex and lagged associations: PM₁₀ exhibits a delayed effect at four months, while NO₂ is associated with immediate increases on mortality. Temperature acts as a nonlinear and oscillatory predictor, with both acute effects and delayed cold-related risks over several months. The findings indicate that pandemic-related mortality disrupted typical seasonal patterns and hindered the detectability of subtler environmental effects. Comparing models that include or exclude COVID-19 mortality reveals clearer environmental associations in the non-COVID model, highlighting the value of distinguishing acute shocks from long-term conditions. Overall, the results underscore the need for stricter air quality standards during colder months and for adaptive public health strategies that consider cumulative and lagged environmental effects. While this study focused on adults aged 30 years and older, the findings suggest that older adults, individuals with chronic respiratory or cardiovascular conditions, and those in disadvantaged settings may be more susceptible to environmental stressors.