A study, conducted by researchers at Mass General Brigham, found a strong link between social risk factors and the development of long COVID. The research, published in Annals of Internal Medicine on 29 July 2025, examined 3700 people who had a SARS-CoV-2 infection during the Omicron variant outbreak between October 2021 and November 2023.
Long COVID Risk Strongly Shaped by Social Determinants of Health According to a Major U.S. Study
Background
The persistence of coronavirus symptoms beyond the acute infection, commonly called long COVID, remains a global health concern. Long COVID, also referred to as post-acute sequelae of SARS-CoV-2 infection, is characterized by symptoms lasting at least three months. These symptoms often include fatigue, breathing difficulties, brain fog, sleep problems, and chronic pain that severely impact daily life.
It is true that biological and clinical factors are central to understanding long COVID. Researchers have also increasingly focused on the role of social and economic conditions. A major investigation supported by the National Institutes of Health through the RECOVER-Adult cohort has now demonstrated that social determinants of health significantly influence the risk of developing this debilitating post-viral condition.
The study analyzed nearly 3700 adults across 33 states, the District of Columbia, and Puerto Rico, all of whom had confirmed SARS-CoV-2 infection between October 2021 and November 2023, during the Omicron wave. Participants completed surveys on personal circumstances shortly after infection, followed by six-month assessments of persistent or newly developed symptoms meeting the long COVID criteria.
Several social determinants of health, such as financial hardship, food insecurity, lower educational attainment, experiences of medical discrimination, lack of social support, and inability to seek medical care due to cost were examined. Neighborhood-level factors such as household crowding, poverty rates, and area-level deprivation were incorporated to capture community conditions influencing recovery trajectories.
Findings
The results were striking. Financial hardship and food insecurity each more than doubled the risk of developing long COVID, with adjusted marginal risk ratios of approximately 2.36. Individuals with less than a college education had about a 60 percent higher risk, while those who skipped medical care due to costs faced almost a threefold greater likelihood of long COVID. Below are the further details of the findings:
• Financial Hardship Doubles Risk: Participants experiencing financial strain at the time of infection had about 2.36 times higher risk of developing long COVID six months later, even after adjusting for illness severity, vaccination, and demographics.
• Food Insecurity Also Doubles Risk: Individuals reporting difficulty accessing adequate food had a similar 2.36 times higher risk of long COVID. This indicates that access to basic resources directly influences COVID recovery prospects.
• Lower Education Increases Vulnerability: Adults without a college degree faced a 60 percent higher risk of long COVID compared with those who attained higher education. Educational attainment appears linked to long-term resilience.
• Skipping Medical Care Triples Risk: Those who avoided or delayed care due to financial barriers were at 2.87 times greater risk of long COVID. This underscores how the absence of affordable healthcare access worsens long-term health outcomes.
• Medical Discrimination Elevates Risk: Participants who experienced discrimination in healthcare settings faced more than double the risk. This shows the damaging health impact of systemic inequities within the medical system.
• Risk From Absence of Social Support: People with weak social support networks had a 1.79 times higher risk of persistent long COVID symptoms. This highlights the protective role of community and social connectedness in recovery.
• Neighborhood Crowding Adds Risk: Living in areas with high household crowding was also associated with increased long COVID risk. This illustrates that the disadvantages at the community level can amplify individual vulnerability.
• Cumulative Effect of Multiple Risk Factors: The risk of long COVID rose progressively with the number of social risk factors a person experienced. This dose-response pattern shows that overlapping vulnerabilities compound long-term risks.
• Consistency Across Different Groups: Racially and ethnically minoritized groups carried more social risk factors overall. But the size of the association between each risk factor and long COVID was similar across racial and ethnic groups.
Implications
Regression models adjusted for confounding variables, including age, sex, comorbidities, acute illness severity, vaccination status, pregnancy, and disability, were used. Controlling for these helped strengthen the evidence that social risk factors independently contribute to long COVID risk beyond clinical and biological characteristics. The models make these findings robust across diverse subgroups and conditions.
The results have important public health and policy implications. They argue that efforts to address long COVID must go beyond clinical care and vaccines to include structural interventions. These should involve improving food security, expanding affordable healthcare, providing financial assistance, combating medical discrimination, and strengthening community networks of social support across vulnerable populations.
However, although the study is among the largest of its kind, it remains observational and cannot prove causation. The authors note potential limitations, including reliance on self-reported surveys, focus on infections during the Omicron period, and a six-month follow-up rather than long-term outcomes. Future research will examine one-year persistence, child populations, and symptom-specific associations with social determinants.
FURTHER READING AND REFERENCE
- Feldman, C. H., Santacroce, L., Bassett, I. V., Thaweethai, T., Alicic, R., Atchley-Challenner, R., Chung, A., Goldberg, M. P., Horowitz, C. R., Jacobson, K. B., Kelly, J. D., Knight, S., Lutrick, K., Mudumbi, P., Parthasarathy, S., Prendergast, H., Quintana, Y., Sharareh, N., Shellito, J., … Karlson, E. W. 2025. “Social Determinants of Health and Risk for Long COVID in the U.S. RECOVER-Adult Cohort.” Annals of Internal Medicine. DOI: 7326/annals-24-01971