Evidence-Based Reviews

Anorexia nervosa and COVID-19

Author and Disclosure Information

Understanding the relationship between eating disorders and infections can improve outcomes.



Recent concerns surrounding coronavirus disease 2019 (COVID-19) make it timely to reexamine the complex findings related to eating disorders and the immune system, and the risks for and detection of infection in patients with anorexia nervosa (AN) and similar disorders. To date, there are no published studies evaluating patients with eating disorders and COVID-19. However, it may be helpful to review the data on the infectious process in this patient population to improve patient communication, enhance surveillance and detection, and possibly reduce morbidity and mortality.

The Centers for Disease Control and Prevention (CDC) and the World Health Organization (WHO) issued warnings that individuals who are older, have underlying medical conditions, and/or are immunocompromised face the greatest risk of serious complications and death as a result of COVID-19, the disease process caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Due to malnutrition, patients with eating disorders, especially AN, may be perceived to have an increased risk of medical conditions and infection. Despite many studies on specific changes and differences in the immune system of patients with eating disorders, the consequences of these changes remain controversial and inconclusive.

This article reviews research on eating disorders, focusing on published data regarding the effects of AN on the immune system, susceptibility to infections, infectious detection, and morbidity. We also discuss clinical considerations related to COVID-19 and patients with AN.

Infection risks: Conflicting data

In a 1981 study that included 9 participants, Golla et al1 concluded that patients with AN may have “resistance” to infections based on a suggested protective factor within the immune system of these patients. Because this study has been cited repeatedly in multiple articles about AN and cell-mediated immunity,2-7 some clinicians have accepted this evidence of resistance to infection in patients with AN, which may lower their suspicion for and detection of infections in patients with AN.

However, studies published both before and after Golla et al1 have shown statistically significant results that contradict those researchers’ conclusion. A study that compared the medical records of 68 patients with AN with those who did not have AN found no significant difference, and concluded that the rate of infection among patients with AN is the same as among controls.8 These researchers noted that infection rates may be higher among patients with later-stage, more severe AN. In a 1986 study of 12 patients with AN, Cason et al9 concluded that while cellular immunity function is abnormal in patients with AN, their results were not compatible with prior studies that suggested AN patients were more resistant to infection.1,2,8

More recently, researchers compared 1,592 patients with eating disorders with 6,368 matched controls; they reviewed prescriptions of antibacterial, antifungal, and antiviral medications as a measure of infection rates.10 Compared with controls, patients with binge eating disorder (BED), patients with bulimia nervosa (BN), and males with AN more often received prescriptions for antimicrobial medications. There was no statistically significant difference between controls and females with AN, which is consistent with other reports of no increased or decreased risk of infection among females with AN. In terms of antiviral use, this study showed an increased prescription of antivirals only in the BN group.

Several other studies examining the rate of infection in patients with AN concluded that there is neither an increased nor decreased rate of infection in patients with AN, and that the rate of infection in this population is similar to that of the general population.8,10-12 Because studies that have included patients with AN have evaluated only symptomatic viral infections, some researchers have proposed that patients with AN may show lower rates of symptomatic viral infection but higher rates of asymptomatic infection, as evidenced by higher viral titers.6 Further research is required. Despite controversy regarding infection rates, studies have found that patients with AN have increased rates of morbidity and mortality from infections.6,12-16

Continue to: Obstacles to detecting infections


Next Article: