COVID-19 and mental health: anxiety and depression on the rise

Younger people and those with lower resilience are more at risk when experiencing high-stress situations. This is what has emerged from the study conducted by the Biomedical Sciences Department of Humanitas University, two years after the first pandemic wave, together with Humanitas’ Center for Anxiety and Panic Disorders San Pio X. The study, whose results were obtained using a Machine Learning approach, was published in The Journal of Neuropsychiatry and Clinical Neurosciences of the American Neuropsychiatric Association.

During the early stages of the pandemic, depressive disorders (8%) and anxiety (11%) affected mostly those with less experience in dealing with stressful events: this is what emerges from the study carried out by Humanitas University and Humanitas’ Center for Anxiety and Panic Disorders San Pio X Hospital published in The Journal of Neuropsychiatry and Clinical Neurosciences (bit.ly/3IxNu41).

The goal of the research, based on Machine Learning techniques, was to identify the most vulnerable categories amongst the population, in order to enhance public intervention campaigns, such as the psychologist bonus 2022, aimed at mitigating the harmful effects of the pandemic on mental health.

The study “Predicting New-Onset Psychiatric Disorders Throughout the COVID-19 Pandemic: A Machine Learning Approach”, led by Prof. Giampaolo Perna and coordinated by Dr. Daniela Caldirola, is based on data from an online survey conducted by researchers among the Italian population during the first two pandemic waves of May-June 2020 and September-October 2020. Of the 3532 participants aged 18 years or older, 736, 80% of whom were women, were eligible to participate because they had never suffered from a mental disorder and had not contracted Covid before the pandemic.

“The data confirm the negative impact the pandemic had on psychological well-being – says Prof. Giampaolo Perna    professor at Humanitas University and Head of Humanitas’ Center for Anxiety and Panic Disorders San Pio X. During the first and second waves, 16% and 18.6% of participants, respectively, developed a mental health disorder, manifesting mainly depressive disorders (8%) and anxiety (11%). In addition, 33% of the respondents reported increased difficulty in managing children in the first wave and 43% in the second wave, as well as work-related fatigue which was reported by 27% and 24% of the respondents during the first and second waves, respectively. Having low resilience, i.e., fewer resources to manage stress, was found to be an underlying factor promoting the development of a mental health disorder in response to the pandemic. High levels of stress due to the risk of spreading the virus, the imposition of restrictive measures, living alone or having stopped engaging in physical activity during the pandemic were also found to be factors leading to possible mental health issues. However, while having contracted Covid and having a history of mental illness are factors that undoubtedly affect mental health in response to the pandemic, never having had a psychiatric disorder and having ‘escaped’ Covid are far from a guarantee for our mental health.”

“The data shows that those who have developed a mental health disorder are about 8-10 years younger than those who have not experienced them – explains Dr. Daniela Caldirola, psychiatrist at Humanitas’ Center for Anxiety and Panic Disorders of San Pio X and researcher at Humanitas University. The average age of participants who developed a new mental disorder was 37.1 years in the first wave and 31 years in the second wave, compared with an average age of 47 and 39.2 years, respectively, among those who did not develop any disorder.”

The support of Machine Learning

The study by Humanitas University and Humanitas San Pio X has identified a Machine Learning predictive model capable of highlighting the most important risk factors for the health and mental well-being of Italians during the pandemic. Thanks to a self-implementing algorithm that learns through data obtained from responses to online questionnaires, the researchers were able to identify multiple factors which could potentially predict the development of a mental health disorder in people without pre-existing mental disorders and previously not affected by Covid. In the initial model, 46 variables were included, mainly focusing on individual pre-pandemic aspects and personal experiences related to the pandemic (including occupation, marital status, medication intake and diseases, habits and lifestyles, child and work management, feelings related to the ability to manage stress, and recovery time). During the Machine Learning training phase, the variables were reduced to the 8 most important ones that then led to the published results.

From results to practice: what to do?

“The results of this study,” emphasizes Prof. Giampaolo Perna,” suggest that reducing stress and anxiety, enhancing resilience, fighting loneliness, and encouraging physical activity and sports might have an important protective role in promoting mental well-being and avoiding the occurrence of psychiatric disorders in response to pandemic stress. Large-scale information and psychological support campaigns during public health emergencies could help those who are most susceptible to stress to better overcome and sustain such challenges.”