We characterized long COVID in a prospective cohort of patients with mild and severe COVID-19 with a systematic clinical follow-up. The prevalence of any symptom at six months was elevated (48%), with a different distribution between hospitalized and non-hospitalized patients. Persisting symptoms did not differ between those patients who needed critical care and those who did not.
Long COVID has been recognized as a public-health problem and a matter of concern for COVID-19 surveys. Several studies have reported information in series of patients in clinical follow-up after the acute infection, but many questions remain unsolved. One of them is if long COVID might be gender-dependent. In our study, females showed a higher prevalence of persisting symptoms in hospitalized and in non-hospitalized groups. Dyspnoea, headache, fatigue and hair loss were all more common in females than males. A prior study performed by Chaoling-Huang et al., also described a higher frequency of long COVID symptoms in females than males [8]. To the best of our knowledge, there are only few studies evaluating long COVID symptoms between males and females, both in hospitalized and non-hospitalized setting [10].
In the recent months, more than 200 persistent symptoms have been described after the acute phase of COVID-19. Most of them are related to the respiratory tract (dyspnoea, cough, chest pain), the main target of SARS-CoV-2. However, many patients report other type of symptoms such as fatigue, cognitive impairment or smell disorders [11].
Chest symptoms were the most common reported affection, with greater prevalence in the hospitalized group. This difference may be related to pneumonia, respiratory distress and lung damage, which were more frequent in hospitalized subjects. The prevalence of chest symptoms in this group raised to 52%, slightly higher than observed in previous reports [9]. However, even in the absence of pneumonia, 15% of non-hospitalized patients reported persisting chest symptoms. Recently, Blomberg et al., described long COVID symptoms in Norwegian non-hospitalized and hospitalized population with a slightly higher prevalence than in our study (52.0% vs. 38.2% respectively) [10].
Fatigue was also frequent, reaching up to 20.9% in the hospitalized group. Although this percentage is relevant, it was lower than others reported before [12, 13], around 65–70%. However, these studies had a shorter follow-up period (90 to 100 days after hospitalization). The reason of fatigue remains unknown, but it is probably a result of a combination of factors, including direct nervous and muscle damage due to SARS-CoV-2 replication [5, 6], immune activation and immune dysregulation, or emotional factors [7].
On the other hand, myalgia and arthralgia were less frequent than in previous studies. Sykes et al., described a prevalence of 50% of these symptoms in hospitalized patients at 100 days after discharge [14].
Overall, persisting anosmia was reported in 6.8% of subjects and was more common in the non-hospitalized group (10.5% vs. 5.2%, respectively). This prevalence is lower than published in previous studies [15, 16]. In addition, prior reports have suggested that anosmia may be associated with a mild to moderate COVID-19 and it could be more common in females than males [15]. However, the low prevalence of anosmia observed in our study cannot confirm this hypothesis.
At six months, 9 subjects reported unspecified neurological symptoms such as poor concentration, loss of recent memory or inability to focus. These unspecified symptoms are named brain fog by some authors, but its origin or relationship with coronavirus disease remains unknown [5, 17]. Brain fog syndrome after COVID-19 has been reported previously [18, 19], although there is not a validated diagnostic criteria established. The studies focused on neurological sequelae of COVID-19 did not use a unanimous criterion and therefore, the conclusions are very heterogeneous. The lack of knowledge about this type of symptoms makes more difficult to study the real prevalence or to establish a relationship with coronavirus disease. Another neurological symptoms like sleep or mood disorders have been reported with a similar prevalence than in our study [20]. However, psychological conditions, stress and isolation may be associated with these symptoms, alongside with coronavirus infection. The higher prevalence on the hospitalized group may be due to the severity of illness, admission to critical care or treatments (e.g., corticosteroids, immunosuppressants).
We identified COPD, female gender, tobacco consumption and the need for hospitalization as predictors for long COVID. Female gender has been reported before as a risk factor for hospitalized subjects but there is a lack of information about other predictors of long COVID.
Our findings suggest that some of the persisting symptoms could be related to a previous lung comorbidity and hospitalization. SARS-CoV-2 damage on respiratory tract could be more intense in patients with previous lung disease, increasing the risk of long-term symptoms. In addition, pneumonia and ARDS may also contribute to develop persisting symptoms.
This study has few limitations. Firstly, some of the symptoms reported are subjective and based on the patient’s testimony (e.g. fatigue, headache, dyspnoea). Additionally, the lack of validated scales to measure most of the symptoms makes difficult to compare data between subjects or studies. Secondly, most of the symptoms may be affected by personal, psychological or environmental factors. We needed to use a non-validated custom questionnaire, as there is not a specific post-COVID-19 document established. Many symptoms could be higher reported in one group than another for unrelated reasons with coronavirus disease. This is the case for hair loss that may be more noticed in women and therefore more reported, leading to a possible increase of the prevalence in female subjects. In addition, previous comorbidities and age could increase the persisting symptoms in the hospitalized group. For COPD patients, we did not record symptoms that started before COVID-19 diagnosis. Therefore, it is possible that COPD patients reported more frequently chest or respiratory symptoms.