Settings and respondents

This case-control study included 878 Bangladeshi adults (cases = 439). The cases were laboratory-confirmed patients who previously experienced COVID-19. The controls comprised individuals without COVID-19 who were matched with cases’ sex and age [16].

Sample size determination

Laboratory-tested 2.0 million COVID-19-positive cases across Bangladesh were considered the population [7]. A 95% confidence level, a 50% response distribution, and a margin with a 5% error were used to ascertain 385 respondents for case data [17].

Inclusion and exclusion criteria

Participants aged > 18 years residing in Bangladesh were included in this study. Patients with acute COVID-19, who were pregnant, and who were unreliable and unpredictable individuals (i.e., participants diagnosed with a severe psychological disorder, bedridden individuals, participants with severe chronic illnesses, particularly rheumatoid and gouty arthritis, cerebrovascular accident, or malignancy) were excluded from this study.

Data source and collection

Six previously trained expert researchers were engaged in the data collection process. Six hundred laboratory-tested post-acute COVID-19-positive individuals were collected from 10 conveniently selected government-affiliated COVID-19 testing centers in Bangladesh. After considering the inclusion and exclusion criteria, 450 participants were interviewed. The aims of this study were explained, documented, and scripted. Written informed consent was obtained from the expected participants to collect, analyze, and publish the data before the interview started. Finally, the participants were interviewed in person using a paper-based semi-structured questionnaire at their residences or occupational places. This study included 439 eligible data points for the “case” group.

Subsequently, data from 439 non-COVID-19 participants were collected. “Control” participants were selected from the case’s suitable family members, neighborhood, or office colleagues.

Therefore, 878 data points were collected between February 24 and April 7, 2022. Optimal confidentiality was maintained throughout the data collection process. Data were recorded anonymously and stored in a password-encrypted computer in an unidentifiable manner for analysis.

Questionnaire

Sociodemographic characteristics including age, biological sex, marital information, education, occupational history, and domestic monthly earnings in Bangladeshi taka (BDT) (1 US dollar = approximately 100 BDT), and present residence were included in the first part of the questionnaire. Phase two of the questionnaire assessed the respondents’ chronic comorbidities, particularly bronchial asthma, hypertension, diabetes mellitus, and renal disease, tobacco use history, and regular physical exercise habits. Dichotomous options (yes or no) were used to answer these questions.

For the case data, the participants provided details about their COVID-19. Symptom severity (very severe, severe, moderate, and mild), and the treatment amenities patients had used (hospital’s intensive care unit, hospital’s general ward, or home) were reported. The recovery duration from the acute COVID-19 vaccination status was also documented for this cohort.

The final part of the questionnaire assessed the pseudoneurological health complaints of the last 30 days using the pseudoneurology sub-scale of Eriksen et al.’s [18] Subjective Health Complaints scale. The PNCs included seven components: palpitation, heat flashes, sleeping problems, fatigue, dizziness, anxiety, and depression. The severity of PNCs was evaluated on a scale of four points (0–3: 0, none; 1, some; 2, much; and 3, severe). These complaints were additionally enumerated based on the number of days in the past month. Severity was multiplied by duration to obtain a complete score ranging from 0 to 90, signifying the degree of illness [18]. For this study, respondents who complained of at least some problems for 3 days (1 × 3 = 3) in the last month and scored ≥ 3 were considered to have PNCs.

Ethical consideration

The ‘Ethical Review Committee (ERC) of Uttara Adhunik Medical College’ cleared the ethical issue for this study (Approval number: UAMC/ERC/Recommend- 11/2021). Furthermore, prospective registration for this case-control study was received from the World Health Organization-endorsed Clinical Trial Registry, India (CTRI/2022/02/040449, registered on February 2, 2022). Formal written informed consent was obtained from all the participants before data collection to collect, analyze, and publish their data. The Strengthening the Reporting of Cohort, Cross-Sectional and Case-Control Studies in Surgery guideline [19] was strictly followed throughout the study.

Participants and public involvement

The respondents and the public were not engaged in our study design, conduct, reporting, and dissemination plans. The study’s endeavors and objectives were elucidated, and assurance of anonymity was granted before receiving written informed consent from the participants.

Analysis of data

The Statistical Package for the Social Sciences (version 28.0; IBM Corp., USA) was used to complete the data analyses. To compute the prevalence, the four responses to the PNC questions were dichotomized according to whether the participants were experiencing the symptoms (yes) or not (no). Chi-squared tests were used to compare categorical variables with and without PNCs. Multiple logistic regression models were used to calculate adjusted odds ratios (aORs), considering a confidence interval (CI) of 95% with PNCs as a measured variable and sociodemographic particularities and impersonal COVID-19 illness-associated components as the predictor variables for PNCs. The two regression models included only the statistically significant variables in the descriptive analyses. The Hosmer–Lemeshow goodness-of-fit test tested the fitness of the model. P-values < 0.05 were considered statistically significant.

General characteristics

This study analyzed 878 data (50.5% women) of patients with a mean age of 38.30 [standard deviation (SD) ± 12.77] years. There were 439 data (49.2% women) in the case group, and the mean age of this group was 38.33 (SD ± 12.53) years. Conversely, the control group comprised 51.7% of women, with a mean age of 38.28 (SD ± 13.01) years.

For all data, most of the participants were married (83.6%), graduates (34.1%), and jobholders (33.7%), had a monthly income of BDT > 45,000 (42.5%), belonged to the nuclear family (66.6%), and were urban dwellers (61.5%). Nonetheless, only 24.0%, 23.5%, 10.3%, and 16.4% of the participants reported hypertension, diabetes, kidney disease, and asthma, respectively. Additionally, approximately 19.8% and 38.8% of the respondents reported that they performed routine physical workouts and were tobacco addicts, respectively. The results are presented in Table 1.

In the case data (Table 2), a more significant part of the attendees was married (85.6%), graduated (33.9%), was jobholder (34.4%), earned BDT > 45,000 (45.1%), belonged to a nuclear family (67.9%), and resided in an urbanized area (68.6%). Nonetheless, approximately 27.3%, 26.7%, 14.4%, and 23.9% of the participants had hypertension, diabetes mellitus, renal illness, and asthmatic problems, respectively. However, approximately 24.4% and 44.9% of the respondents performed routine exercise and used tobacco, respectively. Similarly, a significant proportion of participants with post-acute sequelae COVID-19 recovered from the sickness > 180 days previously (36.7%), showed mild symptoms (49.9%), received treatment at home (65.1%), and received two doses of vaccines before apprehending this disease (53.0%).

Descriptive analysis of the entire data

The prevalence of PNCs was 40.0% in all participants. Nonetheless, the incidence rate of PNCs was considerably higher in the case group than in the control group (67.4% vs. 12.5%, P ≤ 0.001). Closer analysis indicated that the prevalences of PNC components for the case and control groups were as follows: dizziness (17.8% vs. 3.6%, P ≤ 0.001), heat flashes (15.7% vs. 3.2%, P ≤ 0.001), palpitation (35.1% vs. 4.3%, P ≤ 0.001), tiredness (45.1% vs. 12.3%, P ≤ 0.001), sleep problems (48.3% vs. 24.37%, P ≤ 0.001), depression (40.2% vs. 8.7%, P ≤ 0.001), and anxiety (54.7% vs. 11.4%, P ≤ 0.001) (Figure 1). Additionally, participants with lower education (47.2%, P ≤ 0.001), with businesses (48.9%, P ≤ 0.001), who had lower gross monthly earnings (55.0%, P ≤ 0.001), who were city dwellers (43.9%, P = 0.007), who were diagnosed with hypertension (48.8%, P = 0.003), and who consumed tobacco (46.6%, P = 0.001) had significantly higher rate of PNCs (Table 1).

Display full size

Figure 1. 

Distribution of PNCs in controls and cases

Descriptive analysis of case data

Evaluation of case data suggested a higher prevalence of PNCs among business people (73.0%, P ≤ 0.001), respondents with lower income (91.1%, P ≤ 0.001), city dwellers (72.8%, P ≤ 0.001), individuals who recovered from the illness 61–90 days previously (88.2%, P ≤ 0.001), who had moderate symptoms (78.9%, P = 0.033), and who received two doses of vaccine (76.1%, P = 0.010). In contrast, participants diagnosed with kidney disease (44.4%, P ≤ 0.001) and asthma (57.1%, P = 0.010) had lower PNC rates than those without kidney disease and asthma (Table 2).

Regression model of all data

Regression analysis 1 (Table 3) revealed the robust predictability of COVID-19 for PNCs (aOR, 21.97; 95% CI, 14.364–33.604) after adjusting for potential confounders. The highest odds of PNCs were also reported in participants with low income (aOR, 8.558; 95% CI, 4.287–17.087), who were city dwellers (aOR, 1.773; 95% CI, 1.092–2.879), and with hypertension (aOR, 1.903; 95% CI, 1.257–2.879).

Adjusted analysis of case data

Results of regression model 2 are shown in Table 4. In this case, significantly higher odds of PNCs were observed among participants with low income (aOR, 5.269; 95% CI, 1.631–17.023), who were city dwellers (aOR, 3.313; 95% CI, 1.754–5.592), and who recovered from the illness 30–60 days previously (aOR, 4.444; 95% CI, 1.641–12.033) and 121–150 days previously (aOR, 3.150; 95% CI, 1.053–9.419). Nonetheless, lower odds of PNCs were observed among students (aOR, 0.081; 95% CI, 0.018–0.361).

Strengths and limitations

To the best of our knowledge, this is the first case-control study that compared PNCs among patients with post-acute COVID-19 and healthy individuals. Furthermore, this study added COVID-19-related data from a low-resourced country in the literature, which is scarce. However, this study has some limitations. First, this study measured subjective health complaints in pseudoneurology; thus, the prevalence of anxiety and depression should be used cautiously. Second, this study was cross-sectional, which would not enable us to understand causal relations between dependent and independent variables. Finally, it is not impossible that the control data included some asymptomatic COVID-19 patients, which may have influenced the study results. Despite these limitations, this study found some pieces of evidence for future studies on the association between COVID-19 and neuropsychological disorders.

Conclusion

This study demonstrated a remarkably high prevalence of PNCs, thus predicting upcoming neuropsychological health concerns among the community during and after the COVID-19 pandemic. This study also indicated that individuals with lower household income were significant victims of COVID-19. Healthcare amenities must be prepared to address these issues and mitigate the current and upcoming health burden of PNCs. The underprivileged population should be prioritized when discussing health issues among patients with post-acute COVID-19. Additional surveys are required to monitor the health of patients with post-acute COVID-19.