Social status is another key factor affecting survival during a pandemic

Until now, mathematical models used to describe epidemics have not allowed the combined effects of multiple individual characteristics to be studied in an epidemic simulation. Researchers at the HUN-REN Alfréd Rényi Institute of Mathematics have addressed this issue in recent papers published in the leading journals Nature Communications and Science Advances.

During the recent pandemic in Hungary, Dr Márton Karsai (HUN-REN Rényi Institute, CEU), Dr Júlia Koltai (HUN-REN Centre for Social Sciences), and Dr Gergely Röst (University of Szeged) and their teams, as members of the National Laboratory for Health Security, focused on collecting as much data as possible on how people adapted their social interactions to avoid infection.

Their primary aim was to track changes in age-contact matrices used in epidemic models.

These matrices describe the likelihood of an individual of a given age interacting — for example, a retiree in their 70s —interacting with someone from a specific age group, such as their grandchild, on a given day.

Their study, involving PhD student Adriana Manna, found that in addition to age, factors such as income level, education, employment status, and settlement size also influence the number of social contacts an individual has.

For example, the way people adjust the number of their connections differs depending on whether they are employed or currently unemployed, as well as whether they live in a larger city or a smaller rural settlement.

However, until now, no mathematical method has been available to incorporate these observed characteristics into epidemic models, as previous approaches only considered a single factor—age—when describing contact patterns.

The Hungarian researchers, in collaboration with Dr Nicola Perra from Queen Mary University, addressed this mathematical challenge using multidimensional contact matrices. They demonstrated that this multidimensional approach provides a more accurate model of epidemic dynamics and outcomes.

Epidemiological models based on real data enable much more accurate analyses than previous approaches. They serve as a tool for mapping and understanding inequalities in epidemic outcomes across different social groups.

According to the researchers, these findings are not only of scientific value but can also inform public health decisions by improving infection prediction, risk assessment, and the development of intervention plans.

HUN-REN aims to become one of the most effective research organisations in Europe and to contribute significantly to Hungary's social and economic success. The HUN-REN Act, effective from 1 January 2025, provides the necessary framework for HUN-REN to transform into a more responsive, collaborative, performance-oriented, and well-functioning network, ensuring the freedom of scientific research and the continuation of exploratory research, while providing multidisciplinary solutions to complex challenges affecting both national and international communities. Employees within the modernising HUN-REN can look forward to a long-term, consistent, value-driven, and performance-focused funding system, mutually reinforcing structure and operation, competitive salaries, and a predictable research career.

They looked at age first

Dr Márton Karsai told Hungarian news portal Index us about the beginnings, the successes and the future. He said that in their previous study, in 2019, they collected a certain type of data set, with individuals of a certain age, and looked at their likelihood of meeting others. The encounter was then realised,

if they were within 2 metres of each other for at least 15 minutes.

In March 2020, the Hungarian Data Providers' Questionnaire (MASZK) was launched, with more than 250,000 people participating and more than 400,000 anonymous and voluntary responses. A representative telephone survey was also conducted as part of the project developed with Dr Julia Koltai, first with 1,500 and later with 1,000 people over 26 months.

This large-scale, long-term data collection campaign eventually became one of the largest single-country epidemic behaviour questionnaires in the world. As well as respondents, we also focused on who a person met on a given day,

said Karsai.

The researchers asked respondents to estimate the age, education and social class of people they had met the day before. Later, they added more information to the questionnaire, asking questions about people's protection, vaccinations and travel habits to get a more accurate picture of how their behaviour changed during the pandemic.

The first study measured socio-demographic and socio-economic variables from the respondents' point of view and tried to find out whether they correlated with the number of connections. The results showed that, in addition to age, respondents' level of education, employment status, income and housing type had the greatest influence on the maximum number of daily connections.

The second study went further and organised the data into a multi-dimensional mathematical structure, a so-called generalised contact matrix, which not only contained the probability of contact between age groups, but also broke down the age groups into further dimensions,

Karsai explained.

From this data, they were able to estimate the number of contacts a young unemployed adult has with people of a different age group who are employed. "This information, organised in a generalised contact matrix, can be used in modelling,

leading to more accurate epidemiological models.

Neighbourly behaviour is risky

For the different phases of the coronavirus pandemic, the results showed that during the major lockdowns, people living in small rural towns had higher contact rates than those living in large cities.

People living in rural houses with gardens were able to get out and stay in touch with others, such as neighbours, more easily than urban dwellers, who were more affected by the lockdowns,

said Karsai.

It was also found that there were differences between socio-economic groups in the rate of infection with the virus and its consequences. People in higher social classes were the least likely to have reduced contact and were proportionally the most infected.

People from lower socio-economic groups died in greater numbers,

not because of their contact patterns, but because of their health status and poorer access to health care. Dr Karsai said that epidemiologist Dr Beatrix Oroszi and her team looked at the outcome of the pandemic from the perspective of socio-economic groups and came to the same conclusion.

Cover image (for illustration purposes only): Getty Images