A recent multicohort study investigated whether adult obesity increases the risk of severe infections and assessed its contribution to infection-related deaths globally. The research builds on observations from events like the COVID-19 pandemic, where obese patients often experienced higher rates of severity and mortality.
Study Design and Methodology
The prospective multicohort study integrated individual-level data from the Finnish Public Sector study and the Health and Social Support study in Finland, along with replication data from the UK Biobank. Body mass index (BMI) data, both self-reported and measured, were collected from 1998 to 2010. Participants were categorized into healthy weight, overweight, and three classes of obesity (class I, II, III). Waist circumference and waist-to-height ratio were also evaluated. Individuals with prior severe infections were excluded.
Follow-up involved national hospitalization and mortality registries to identify first incident severe infections, defined as hospital-treated or fatal cases. Statistical models were adjusted for various factors including age, sex, socioeconomic status, lifestyle choices (smoking, alcohol consumption, physical activity), and pre-existing medical conditions.
The hazard ratios derived were then combined with global obesity prevalence and mortality data from the Global Burden of Disease, Injuries, and Risk Factors (GBD) Study to calculate the population attributable fractions for infection-related deaths worldwide.
Key Findings
Data from 67,766 adults in Finnish cohorts and 479,498 adults in the UK Biobank revealed a clear dose-response relationship: individuals with class III obesity had nearly three times the risk of infection-related hospitalization or death compared to those with a healthy weight. All obesity classes were linked to approximately 30%–40% higher overall severe infection risk, with substantially higher risks observed in the most severe obesity categories.
This association remained strong even after controlling for lifestyle, socioeconomic status, and health differences. A similar increase in risk was observed when obesity was assessed using waist circumference and waist-to-height ratio, indicating consistent associations across various adiposity measures. Obesity was found to increase the risk across nearly all infection categories, with particularly strong associations for viral infections (especially acute viral infections) and bacterial infections. Skin and soft tissue infections showed the strongest association, with almost a threefold increase in risk. For COVID-19, a hazard ratio of 2.3 was observed in the UK Biobank. No increased risk was identified for human immunodeficiency virus or tuberculosis.
Weight change analyses further supported these findings; individuals who gained weight from overweight to obesity experienced increased infection risk, while those who lost weight from obesity to overweight or healthy weight showed modest risk reductions.
Globally, approximately 8.6% of infection-related deaths in 2018, 15.0% in 2021, and 10.8% in 2023 were attributable to adult obesity. This translates to about 0.6 million infection-related deaths worldwide in 2023 being linked to obesity.
Public Health Implications
Adult obesity is identified as a significant and consistent risk factor for severe infections across diverse pathogens and populations, with the risk progressively increasing with higher obesity classes. The study suggests that roughly one in ten infection-related deaths globally is linked to adult obesity.
Consequently, prevention of obesity, implementation of evidence-based weight-management programs, and development of vaccination and infection-control policies that account for obesity are deemed critical public health priorities.
Addressing obesity could help reduce hospitalizations, deaths, and healthcare burdens during routine infectious seasons and future pandemics.
Study Limitations
As an observational study, the findings do not prove causality. Global estimates rely on modeling assumptions that may vary across regions. The cohorts used are not fully population-representative, and the possibility of residual confounding cannot be entirely excluded, which may mean absolute risk estimates could differ in other populations.