Did the pandemic change lifestyle behaviours in Italy? An interrupted time series analysis on the four main NCDs behavioural risk factors from 2008 to 2023.

Abstract

Background: The COVID-19 pandemic has had repercussions in several areas. The indirect effects of the pandemic on healthy living behaviours are multiple and complex to assess. The aim is to assess the impact of the outbreak of COVID-19 pandemic in Italy on the most relevant modifiable behaviours risks for non-communicable diseases (NCDs).

Methods: PASSI 2008-2023 data referring to a sample of 18-69-year-olds residing in Italy was used to estimate the prevalences of smoking, alcohol, fruit and vegetable consumption, and physical inactivity lifestyle. For each of these risks was used an interrupted time series (ITS) study with Generalized Least Squares (GLS) model to assess trends before and after the pandemic's outbreak in Italy (March 2020). The "intervention" period is postulated as March 2020, and the "post-intervention" period extends from April 2020 to December 2023. In these models, the hypothetical situation without "intervention" and with the trend remains unchanged is commonly known as the 'counterfactual' scenario. Through ITS model both "counterfactual data" and "factual data" were obtained.

Results: From 2008 to 2023, 532,115 people were interviewed. Results showed, during "post-intervention" period, significant differences between factual and counterfactual prevalences started in 2022 and strengthened during 2023 and for all of the four behavioural indicators analysed: smoking, high-risk alcohol consumption, fruit and vegetable consumption (both all three of them worsening) and physical inactivity (which appears to decrease). Current smokers exhibited similar prevalences in March 2020 (24.5% vs. 24.3%), followed by a plateau throughout the entire 'post-intervention' period. By December 2023, however, the counterfactual prevalence was significantly lower than the observed rate (24.5% factual vs. 22.7% counterfactual). The initial improvement observed in the prevalence of high-risk drinkers in March 2020 (13% factual vs. 17% counterfactual) was followed by a rapid worsening and in December 2023 the prevalence was significantly higher than expected (19.7% vs. 16.9%). The fruit and vegetable consumption worsened and the factual prevalence of five-a-day in December 2023 was significantly lower than counterfactual (6.6% vs. 9.0%). Physical inactivity following an initial worsening in March 2020 (albeit not statistically significant), appeared to decrease in December 2023, with the factual prevalence lower than counterfactual (26.9% vs. 32.4%).

Conclusions: Results found in this article showed at the end of studied period the achievement of a plateau for current smokers, a worsening in the high-risk alcohol consumption, in fruit and vegetable intakes and a reduction in physical inactivity. Understanding shifts in these lifestyle indicators is crucial for the proper design of interventions aimed at reducing the burden of NCDs.