Herbal Medicines for Long COVID: A Phase 2 Pilot Clinical Study

1. Introduction

Although a significant number of patients have reported persistent symptoms following acute coronavirus disease 2019 (COVID-19), a global consensus on terminologies and definitions has yet to be established. The definition of long COVID, also known as chronic or postacute sequelae of SARS-CoV-2 infection (PASC), has, is not well defined, and different countries and institutions use different terms or definitions. According to the World Health Organization, post-COVID-19 conditions are defined as symptoms that usually occur within three months of the onset of COVID-19 symptoms, persist for at least two months, and are not explained by other alternative diagnoses [1]. In South Korea, chronic COVID-19 syndrome (long COVID) is defined as the persistence of one or more symptoms or signs that cannot be explained by another disease after 12 weeks of COVID-19 diagnosis [2]. These symptoms are diverse and can affect various systems, commonly including fatigue, shortness of breath, cough, chest pain, joint pain, muscle aches, “brain fog” (cognitive difficulties), headache, heart palpitations, sleep disturbances, depression, anxiety, and loss of taste or smell. A meta-analysis of 41 studies found that the global combined prevalence of PASC was 43%, with fatigue, memory problems, and dyspnea being the most common symptoms [3]. Similarly, other studies have found that fatigue and cognitive impairment are the most common and frequently reported PASC symptoms [4,5]. In addition, unlike other common symptoms of PASC, such as dyspnea and depression, there are currently no established and effective treatments for these conditions [6]. It is important to note that the prevalence of long COVID can vary based on factors such as the definition of long COVID used in the study, the population being studied, and the follow-up duration.

Several risk factors appear to be associated with PASC, including female gender, ethnic minority status, socioeconomic deprivation, smoking, obesity, and comorbidities [7]. Data from the United States, 153,543 adults, including 19,985 with long COVID, revealed that 14% of adults aged 18–84 (35.11 million) and 15.5% of adults at the working age of 18–64 (30.65 million) had long COVID by November 2022 [8]. In addition, approximately 27.7 million adults and 24.2 million adults at the working age may experience adverse socioeconomic and mental health outcomes due to long COVID. Furthermore, the estimated annual earnings loss among working-age adults was $175 billion. These findings highlight the significant public health and economic impact of long COVID, underscoring the need for further investigation and intervention [8].

Despite the significant impacts of long COVID, there is currently a lack of treatment methods supported by concrete evidence. There are several approaches available to manage symptoms and improve overall well-being in individuals with long COVID. For fatigue in patients with long COVID-19, treatment options include medications, alternative medicine, cognitive behavioral therapy, and exercise therapy. However, studies evaluating the effectiveness of these interventions are limited and lack high-quality evidence [9]. To date, the treatment strategies for various post-COVID-19 syndromes are inadequate. However, traditional medicine has been identified as a potential solution due to its ability to relieve symptoms, improve organic injuries, and affect immune function [10]. According to a previous study, a significant number of patients with post-COVID-19 syndrome who were identified with particular syndromes, such as Lung and Spleen Qi Deficiency and Qi and Yin Deficiency, in traditional East Asian medicine showed improvement in symptoms after treatment [11]. Furthermore, randomized controlled trials have suggested that traditional medicine can improve fatigue in post-COVID-19 patients [12]. However, these pieces of evidence appear to be insufficient for complementary and alternative medicine (CAM). There is a lack of published studies regarding the effectiveness and safety of CAM interventions for long COVID [13], highlighting the need for more clinical studies in this area.

According to reports, Korean medicine (KM) doctors consider Qi Deficiency as a major cause of chronic fatigue syndrome and prioritize prescribing Bojungikki-tang (BIT; Bu-Zhong-Yi-Qi-Tang in Chinese or Hochu-ekki-to in Japanese) as a treatment [14]. In Japan, Bojungikgi-tang is also frequently used to treat fatigue associated with long COVID [15]. During the pandemic period, the most commonly prescribed herbal medicine in South Korea for fatigue, weakness, and dry cough was Kyungok-go (KOG) [16]. Furthermore, Cheonwangbosim-dan (CBD) has the potential to improve or treat physical and mental illnesses, such as cognitive dysfunction, neurosis, insomnia, and cardiac malfunction-induced disease [17]. Therefore, these herbal medicines are considered potentially useful in alleviating long COVID symptoms. It is necessary to establish clinical evidence for these herbal medicines, not only to develop traditional medical treatment strategies for long COVID but also to provide them to the growing number of patients.

This was a phase 2 preliminary clinical trial that aimed to evaluate the safety and efficacy of frequently used herbal medicines in patients experiencing fatigue or “brain fog” associated with long COVID. The research objectives were to assess the feasibility of the intervention and gather evidence for future large-scale studies. Our data demonstrate a new possibility of herbal medicines against long COVID symptoms.

2. Materials and methods

2.1. Trial design

This was a prospective pilot study conducted at Kyung Hee University Korean Medicine Hospital in Seoul, Republic of Korea, with the aim of testing the feasibility and acceptability of using selected herbal medicines in patients experiencing fatigue or “brain fog” after COVID-19. This study was approved by the Institutional Review Board of Kyung Hee University KM Hospital (KOMCIRB 2020-12-002-001). The protocol of this study was published as a protocol paper [18] and registered at cris.nih.go.kr (KCT0006252) prior to the start of the clinical trial.

2.2. Participants

Patients with fatigue or cognitive symptoms were recruited for this study. Each patient was prescribed three KM herbal medications based on syndrome differentiation and symptom type. Since this clinical trial did not include a control group, randomization, allocation concealment, and blinding were not implemented. Patients had monthly visits for a 12-week medication period, and their symptoms and blood biochemistry were assessed up to 36 weeks after participation. The enrollment, intervention, and assessment schedules are summarized in Table 1.

This study included individuals over the age of 19 who had recovered from COVID-19, reported persistent fatigue or cognitive dysfunction for more than four weeks after their diagnosis but not before, achieved a total score of more than 76 points on the Checklist Individual Strength (CIS), had no overall cognitive function issues, and could provide written informed consent. However, individuals with conditions that could potentially cause fatigue or “brain fog” or interfere with drug intake or absorption as well as those with other allergic, hepatic, or nephrotic diseases, were excluded from the study. All participants provided written informed consent prior to participating in this study. All of the entry criteria are listed in Table S1.

2.3. Intervention

The KM syndrome differentiation was conducted by a Korean Medicine doctor specializing in internal medicine specialist, with over 10 years of clinical experience. Patients were classified into either the fatigue or cognitive dysfunction group based on the primary symptoms observed after COVID-19. Subsequently, they were assigned to the appropriate herbal medicine intervention group according to their specific sub-syndrome differentiation pattern. The criteria for syndrome differentiation can be found in Table 2. Based on syndrome differentiation, three different herbal medications were prescribed to each patient for 12 weeks. Among patients with fatigue, the pattern/syndrome of Lung-Spleen Qi Deficiency group received BIT (Kracie Bojungikgitang Extract Fine Granule, Kolmar Pharma, Korea) as one sachet (3.75 g) twice a day before or in between meals. In addition, the syndrome of Dual Deficiency of Qi and Yin group received KOG, KBPharm, Korea) as one sachet (20 g) twice a day in the morning and evening before or in between meals. Furthermore, the cognitive dysfunction group received CBD (Soonsimhwan, Hanpoong Pharm, Korea) as one sachet (one dose) once a day in between meals. All medications and their respective dosages were approved by the Ministry of Food and Drug Safety (MFDS). All medication ingredients are detailed in Table 3.

2.4. Outcome measures

The primary endpoint of this study was to determine the treatment success rate after 12 weeks of herbal medication for fatigue and cognitive dysfunction in patients who had recovered from COVID-19. The fatigue or cognitive dysfunction was assessed before and 13 weeks after the administration of herbal medicines using a Visual Analog Scale (VAS) score ranging from 0 to 100. A change in VAS score between before and after treatment greater than 15 points was defined as treatment success. The secondary endpoints included evaluating medication adherence and utilizing CIS [19], Chalder Fatigue Scale and Subscale [20], EQ-5D-5L for health-related quality of life [21], Pittsburgh Sleep Quality Index-K [22], Korean–Montreal Cognitive Assessment scale for cognitive function [23], Cognitive Failure Questionnaire for cognitive function [24], Beck’s Depression Inventory [25], Digit Span Test in the Korean–Wechsler Adult Intelligence Scale for assessing changes in cognitive dysfunction [26], and Korean–Boston Naming Test-15 for measuring language proficiency in cognitive impairments [27]. During the 12-week treatment period, patients had monthly visits. Following that, they had follow-up visits every 12 weeks over a total period of 24 weeks for outcome assessment.

To evaluate the study’s feasibility, the patient recruitment rate was calculated as the ratio of the number of recruited patients during the study period to the planned number of study patients (45 and 15 patients per intervention group, respectively). Dropout reasons were documented, and the dropout rate was calculated and compared for all study patients as well as for each intervention group.

The safety of the drugs was evaluated by monitoring the incidence of adverse events (AEs). AEs referred to all harmful and unintended symptoms of signs, including abnormal laboratory test results in patients received herbal medicine, without necessarily having a causal relationship. At Weeks 0 and 13, laboratory investigations, including blood chemistry, were performed, and an electrocardiogram was assessed. Laboratory tests determined the levels of red blood cells, hemoglobin, hematocrit, platelets, white blood cells, fasting blood sugar, BUN, creatinine, AST, ALT, and blood electrolytes (Na, K, Cl). AEs were assessed at all in-person visits occurring at 1, 5, 9, 13, 25, and 37 weeks. Detailed research methodology can be found in the previously published protocol paper [18].

2.5. Statistical analysis

To assess the feasibility of the study design, the subject recruitment rate, dropout rate, and dropout reasons were analyzed. Since this trial was conducted as a pilot study, the sample size was not calculated beforehand. The dropout rate was calculated for all participants and for each herbal medication group, and the dropout reasons were also assessed. For continuous variables used to assess the potential clinical effectiveness of the herbal medications, descriptive statistics such as standard deviation, median, minimum, and maximum values were presented for each time point. Statistical analysis was conducted on an intention-to-treat basis, and the last observation carried forward analysis was employed to handle missing data while evaluating effectiveness. An analysis of covariance (ANCOVA) was used to compare the different herbal medication groups (p < 0.05). In cases where the basic assumptions of ANCOVA, such as normality and equal variance of data, were violated, the Kruskal–Wallis test was used. The statistical analysis was conducted using the R software (R 4.1.2, The R Foundation, www.r-project.org, accessed on August 9, 2022). A full analysis set was used to analyze variables for effectiveness evaluation, while data for safety evaluation were analyzed using the safety set. All statistical tests, unless otherwise specified, were two-sided with a significance level of 5%.

3. Results

3.1. Participants

A total of 49 potential participants were screened, with 45 included in the study. A flow chart of patients enrolled in this study is depicted in Figure 1 and Figure 2.

3.2. Baseline characteristics

In this study, fatigue and “brain fog” symptoms were assessed in patients with long COVID. Herbal drugs for fatigue symptoms (BIT and KOG) and “brain fog” (KOG) were prescribed to patients based on their symptoms and pattern identification diagnosis. Patients in the CBD group were older (average age: 51.33 years) than those in the fatigue herbal drug groups (BIT and KOG). The average duration of symptoms was 89.47 days in the BIT group, 114.8 days in the KOG group, and 119 days in the CBD group. The body mass index, educational level, and occupational status were comparable among the groups. When the National Early Warning Score (NEWS) was considered, the KOG and CBD groups showed a more severe status during the acute COVID-19, with fatigue symptoms being more severe in the BIT and KOG groups and “brain fog” symptoms being more severe in the CBD group (Table 4).

3.3. Feasibility assessment

Between August 2021 and July 2022, the targeted recruitment of 45 participants was successfully completed, achieving a 100% recruitment and enrollment rate. Among the included participants, 14 (31%) dropped out due to various reasons. A significant number of these dropouts were attributed to unexpected diseases, with six individuals experiencing a reinfection of COVID-19. However, no withdrawals were linked to AEs from the administered drug. In the BIT group, seven participants dropped out due to unexpected diseases (n = 3), pregnancy (n = 1), and withdrawal of consent (n = 3). In the KOG group, two dropped out due to unexpected diseases. In the CBD group, five dropped out due to unexpected diseases (n = 2) and withdrawal (n = 3) (Table 4 and Figure 1). Medication adherence among participants who completed 12 weeks of drug administration was high (92.27%, 96.26%, and 85.66% in the BIT, KOG, and CBD groups, respectively, Table 5). Considering these factors as well as participant recruitment, intervention administration, and long-term follow-up, the study design can be considered feasible.

3.4. Assessments of fatigue and “brain fog” symptoms

After 12 weeks of herbal drug administration, fatigue symptoms improved significantly in each group, with treatment success rates of 80%, 53.33%, and 46.67% in the BIT, KOG, and CBD groups, respectively. However, treatment success rates for “brain fog” symptoms were comparatively low (40%, 46.67%, and 13.33% in the BIT, KOG, and CBD groups, respectively). In addition, CIS showed significant improvement in each group at 13 weeks (mean difference: −29.53, 95% confidence interval [CI]: −40.96 to −18.2 in the BIT group, −31.47, 95% CI: −43.47 to −19.46 in the KOG group, and −23.4, 95% CI: −34.79 to −12.01 in the CBD group), 25 weeks, and 37 weeks, with no significant difference between groups after treatments. Similarly, the VAS score for fatigue showed significant improvement at 13 weeks (mean difference: −32.53, 95% CI: −46.07 to −19.00 in the BIT group, −30.13, 95% CI: −47.11 to −13.16 in the KOG group, and −16.73, 95% CI: −27.48 to −5.98 in the CBD group), 25 weeks, and 37 weeks in each group, with no significant difference between groups after treatments. In terms of the total, physical health, and mental health scores of ChFQ, significant improvement in each group was observed at 13 weeks (mean difference: −23.67, 95% CI: −35.09 to −12.24 in the BIT group, −21.27, 95% CI: −32.91 to −9.62 in the KOG group, and −10.93, 95% CI: −19.93 to −1.93 in the CBD group for the total score; −17.27, 95% CI: −25.58 to −9.15 in the BIT group, −14.2, 95% CI: −21.41 to −6.99 in the KOG group, and −6.27, 95% CI: −11.99 to −0.54 in the CBD group for the physical health score; −6.4, 95% CI: −10.98 to −1.82 in the BIT group, −7.07, 95% CI: −11.98 to −2.16 in the KOG group, and −4.67, 95% CI: −8.3 to −1.03 in the CBD group for the mental health score), 25 weeks, and 37 weeks, with no significant difference between groups (Table 5).

In terms of “brain fog” symptoms, no significant improvements were observed in each group. The VAS score for “brain fog” did not show significant improvement at 13 weeks (mean difference: −6.27, 95% CI: −26.71 to 14.17 in the BIT group, −5.4, 95% CI: −27.98, 17.18 in the KOG group, and 1.4, 95% CI: −10.8 to 13.6 in the CBD group), 25 weeks, or 37 weeks, with no significant difference between groups. However, the Korean–Montreal Cognitive Assessment score showed significant improvement in the KOG group at 13 weeks (mean difference: 1.80, 95% CI: 0.7 to 2.83), 25 weeks (mean difference: 1.87, 95% CI: 1.03 to 2.70), and 37 weeks (mean difference: 1.47, 95% CI: 0.53 to 2.40). In addition, the Cognitive Failure Questionnaire score showed significant improvement in the BIT and KOG groups at 37 weeks (mean difference: −4.40, 95% CI: −7.57 to −1.23 in the BIT group, −3.93, 95% CI: −11.57 to 3.70 in the KOG group, and −11.67, 95% CI: −21.07 to −2.26 in the CBD group), with no significant difference between groups. Furthermore, there was no significant improvement in the BIT and CBD groups at each follow-up period. Scores for the DF-forward test and DF-backward test did not show significant improvements at 13 weeks (mean difference: 0.27, 95% CI: −0.27 to 0.8 in the BIT group, 0, 95% CI: −0.47 to 0.47 in the KOG group, and −0.07, 95% CI: −0.46 to 0.42 in the CBD group for the DF-forward score and 1.4, 95% CI: 0.4 to 2.4 in the BIT group, 0.4, 95% CI: −0.26 to 1.06 in the KOG group, and 0.8, 95% CI: −0.21 to 1.81 in the CBD group for the DF-backward score), 25 weeks, or 37 weeks. Furthermore, the Korean–Boston Naming Test–15 score did not show significant improvement at 13 weeks (mean difference: 0.13, 95% CI: −0.06 to 0.33 in the BIT group, 0.13, 95% CI: −0.06 to 0.33 in the KOG group, and 0.33, 95% CI: −0.21 to 0.87 in the CBD group), 25 weeks, or 37 weeks (Table 5).

3.5. Other outcome and safety assessments

The EQ-5D did not show a significant difference between groups at 13 and 37 weeks, but there was a significant difference at 25 weeks (mean difference: 0.06, 95% CI: 0.02 to 0.11 in the BIT group, −0.01, 95% CI: −0.05 to 0.03 in the KOG group, and 0.03, 95% CI: −0.06 to 0.11 in the CBD group). In addition, the Pittsburgh Sleep Quality Index-K score showed significant improvement in the KOG group at 13 weeks (mean difference: −2.71, 95% CI: −3.76 to −1.67), 25 weeks (−2.08, 95% CI: −3.28 to −0.89), and 37 weeks (−2.82, 95% CI: −3.94 to −1.70), with no significant improvements in the BIT and KOG groups. Furthermore, the Beck’s Depression Inventory score showed significant improvement in the BIT and KOG groups at 13 weeks (mean difference: −3.80, 95% CI: −6.86 to −0.74 in the BIT group, −4.73, 95% CI: −6.87 to −2.59 in the KOG group, and −5.53, 95% CI: −11.21 to 0.14 in the CBD group), 25 weeks, and 37 weeks (Table 5).

Among the 45 participants, 10 (22.22%) reported AEs, with no significant differences in the incidence rates of AEs among groups (p = 0.2805). All AEs were mild, and none were found to be causally related to the medication. Furthermore, there were no significant differences among groups in terms of the proportion of normal laboratory test results after treatment termination (Table 6).

4. Discussion

Through this pilot study, we found that the research design and intervention for this study were feasible. Although a considerable number of participants dropped out during the follow-up period due to the COVID-19 pandemic, there were no AEs directly related to the intervention, and adherence to the medication was high. Fatigue symptoms improved significantly in all groups, regardless of the type of prescription administered. However, “brain fog” symptoms did not improve significantly, even in the CBD group where this symptom was predominant.

In this study, three interventions were selected as treatments for fatigue and “brain fog” symptoms that may occur after COVID-19. BIT had been widely used as a prescription to treat various symptoms of impaired gastric function caused by dietary irregularities or overexertion. It has been approved by the Korean regulatory agency (Ministry of Food and Drug Safety, MFDS) for weakness, fatigue, lethargy, post-disease debility, poor appetite, night sweating, and vulnerability to fatigue due to a lack of vitality and sluggish gastrointestinal movement. It is composed of Ginseng Radix, Atractylodis Rhizoma Alba, Astragali Radix, Angelicae Gigantis Radix, Zizyphi Fructus, Bupleuri Radix, Citri Unshius Pericarpium, Glycyrrhizae Radix et Rhizoma, Cimicifugae Rhizoma, and Zingiberis Rhizoma Recens. It has been to have many clinical effects, and the therapeutic mechanism may be related to immunological, neuropsychiatric, anti-inflammatory, and musculoskeletal regulation [28]. It has been suggested to be effective in alleviating chronic fatigue-related general symptoms [29]. In addition, it can improve the quality of life, fatigue, and immunological status of elderly patients with weakness [30]. A previous case report reported the use of BIT for fatigue symptoms after COVID-19 [31]. KOG has been widely used to treat chronic diseases and is approved for conditions such as illness-related weakness, post-disease debility, weak constitution, physical fatigue, lethargy, and menopausal disorders, serving as a tonic and strengthener in Korea. It is composed of Rehmanniae Radix, Ginseng Radix, Wolfiporia Cocos, and Honey. Its therapeutic mechanism may be related to immune system activation [32], fatigue reduction [33,34], and cognitive function enhancement [35] as well as its anti-inflammatory effects [36,37,38]. CBD has been used for the treatment of anxiety, palpitation, and cognitive dysfunction. It has been approved in Korea for a variety of conditions, including insomnia, anxiety, restlessness, thirst, palpitations, shortness of breath, nervous debility, “brain fog,” and feverishness. Its therapeutic effect may be attributed to its antioxidant, anticancer, anti-inflammatory, and growth-promoting properties without significant toxicity or AEs [39]. In animal models, it has been found to inhibit learning and memory impairment, suggesting a potential effect on cognitive function [40]. CBD has also been shown to reduce inflammatory responses [41] and improve sleep quality in patients with primary insomnia [42]. Furthermore, it exhibits neuroprotective [43], vasorelaxant, hypotensive [44], and anti-Alzheimer’s disease [45,46] effects. It is commonly prescribed for alleviating the psychological and behavioral symptoms of dementia related to Alzheimer’s disease [47].

This study has some limitations. First, it was a pilot study with no control group. Therefore, concrete evidence regarding the efficacy and safety of these three herbal medications for long COVID could not be established. However, the feasibility of the study design was confirmed, and future large-scale clinical trials are required. Second, the outcomes assessed may be inappropriate for assessing fatigue and “brain fog” in patients with long COVID. The definition of long COVID is currently ambiguous, and there is no clear description of the underlying pathologic mechanism of its symptoms. While the diagnosis and treatment of long COVID-related fatigue may be approached by considering post-viral syndrome, approaching “brain fog” is more challenging due to its distinct presentation from cognitive impairments observed in dementia. One possible reason for the lack of significant improvement in cognitive function before and after drug treatment in patients with “brain fog” in this study is that the condition they are experiencing differs from cognitive decline in Alzheimer’s dementia. As a result, existing assessment tools designed for patients with dementia may not be able to adequately evaluate their cognitive function, which is a significant consideration in this context. Third, approximately 30% of study participants dropped out, demonstrating a high dropout rate. This may be attributed to the study’s long-term participation requirements, which included a 12-week intervention period followed by a six-month follow-up. Furthermore, the ongoing COVID-19 pandemic has led to a significant number of patient re-infections. When conducting clinical trials for the management of patients with long COVID, it is important to consider the potential for a high dropout rate in the study design.

5. Conclusions

The feasibility of this study, which included patients with long COVID experiencing fatigue and “brain fog” symptoms and given three different herbal medicines (BIT, KOG, and CBD) for 12 weeks based on their symptom patterns, has been demonstrated. Given the impact of long COVID and the scarcity of studies evaluating the effects of interventions on fatigue and “brain fog,” this study is deemed valuable. Future clinical studies with an appropriate control group and a sufficient sample size are needed to establish the evidence for herbal medicine in the management of long COVID-19.