Hospitalized or Not, COVID Symptoms Persist in Many After 2 Years

Over half of patients were still experiencing at least one post-COVID-19 symptom 2 years after acute infection, whether they had been hospitalized or not, a cross-sectional cohort study showed.

Among nearly 700 patients infected during the first wave of the pandemic, 59.7% of those who were hospitalized and 67.5% of those who were not hospitalized still had at least one symptom 2 years later (P=0.01), reported César Fernández-de-las-Peñas, PT, PhD, of Universidad Rey Juan Carlos, in Madrid, Spain, and colleagues in JAMA Network Open.

The most prevalent symptoms for hospitalized and nonhospitalized groups at 2-year follow-up were:

• Fatigue: 44.7% vs 47.7%
• Pain (including headache): 35.8% vs 29.9%
• Memory loss: 20% vs 15.9%

“Our results revealed similar proportions of hospitalized and nonhospitalized patients with post-COVID-19 symptoms 2 years after the acute infection, suggesting that, despite having not been hospitalized during the acute phase, the symptoms of long COVID are also found in the nonhospitalized cohort,” the authors wrote. “This finding could be explained by the fact that COVID-19 severity is not a risk factor for the development of long COVID symptoms.”

“Long COVID will require specific management attention independently of whether the patient has been hospitalized or not,” they concluded.

Of note, uninfected controls were not included in this study. “Lack of inclusion of uninfected controls limits the ability to evaluate a direct association of SARS-CoV-2 infection with overall and specific post-COVID-19 symptoms 2 years later,” they wrote. “Accordingly, future studies could include uninfected control populations.”

When patients first presented with COVID-19, the most frequent symptoms were fever, dyspnea, myalgia, and cough, but dyspnea was more prevalent among hospitalized patients (31.1% vs 11.7% of nonhospitalized patients, P<0.001). Anosmia was more prevalent among nonhospitalized patients (21.4% vs 10.0%, P=0.003).

“These differences could be explained by the fact that individuals experiencing less bothersome and less severe symptoms (e.g., anosmia, ageusia, and throat pain) did not seek hospitalization during the first wave of the pandemic,” Fernández-de-las-Peñas and team wrote.

Among hospitalized patients, the number of pre-existing comorbidities was associated with post-COVID fatigue (OR 1.93, 95% CI 1.09-3.42, P=0.02) and dyspnea (OR 1.91, 95% CI 1.04-3.48, P=0.03), while the number of pre-existing comorbidities (OR 3.75, 95% CI 1.67-8.42, P=0.001) and the number of symptoms at illness onset (OR 3.84, 95% CI 1.33-11.05, P=0.01) were linked with post-COVID fatigue among nonhospitalized patients.

Most previous studies looking at post-COVID symptoms had shorter follow-up periods. One meta-analysis that included 40 studies and followed patients for up to 120 days suggested a greater prevalence of post-COVID symptoms among hospitalized patients versus nonhospitalized patients. “Data on nonhospitalized patients are based on follow-up periods no longer than 6 months; accordingly, we cannot directly compare our results with previous data,” the authors noted.

To evaluate COVID-19 symptoms 2 years after infection, Fernández-de-las-Peñas and team included 360 hospitalized patients (mean age 60.7, 45% women) and 308 nonhospitalized patients (mean age 56.7, 59.4% women) from two urban hospitals and several general practitioner centers who were infected with SARS-CoV-2 from March 20 to April 30, 2020. These patients did not experience reinfection over the 2 years of follow-up.

Common comorbidities among hospitalized and nonhospitalized patients included hypertension (33.3% vs 24.7%), diabetes (13.6% vs 4.9%), cardiac disease (11.9% vs 11.0%), and obesity (7.8% vs 10.1%).

Participants were scheduled for a telephone interview 2 years after acute infection. Hospitalization and clinical data were collected from medical records.

Other study limitations besides not including uninfected controls included the fact that the researchers did not control for vaccination status. Furthermore, data were self-reported through telephone interviews, which can lead to recall bias.