Abstract
Background
Post-COVID-19 conditions (PCC) or Long COVID, will linger due to continued circulation of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Understanding the burden and risk factors of PCC could guide development of management guidelines for affected persons.
Methods
Using Population-Based Infectious Disease Surveillance platforms established in Nairobi and Siaya Counties in Kenya, we followed up participants previously infected with SARS-CoV-2 between 01/05/2020 and 30/09/2021 to evaluate the presence and risk factors for PCC. Interviews were conducted from 13/10/2021 to 22/11/2021 to elicit information on the presence of four primary outcome categories: (i) presence of respiratory symptoms, (ii) self-reported non-recovery after SARS-CoV-2 infection, (iii) psychological distress, and (iv) worsening disability. The latter two were evaluated for persons ≥ 18 years old. Risk factors assessed included participants’ demographic and clinical characteristics. Logistic regression models were developed for each outcome adjusted for household-level clustering.
Results
Characteristics of the 832 participants from both sites were as follows; 82.7% were < 50 years, 59.3% were female, 5/511 (1.0%) were vaccinated with ≥ 1 dose of COVID vaccine. For the outcomes, 174/832 (20.9%) had respiratory symptoms, 165/793 (20.8%) reported non-recovery following SARS-CoV-2 infection, 152/511 (29.7%) had psychological distress, while 112/511 (21.9%) had a worsening disability. Females had greater odds of reported non-recovery from COVID-19 illness than males, adjusted odds ratio (aOR) of 1.47 (95%CI, 1.01–2.13). Underlying medical conditions was a significant risk factor for all outcomes: for presence of respiratory symptoms, aOR = 1.82 (95% CI, 1.16–2.87), for reported non-recovery, aOR = 1.93 (95% CI, 1.24–3.02), for psychological distress aOR = 1.87 (95%CI, 1.17–2.99), while for worsening disability aOR = 2.58 (95% CI, 1.54–4.34). Other significant predictors included living in the Asembo site associated with psychological distress (aOR = 2.23; 95% CI, 1.42–3.53), worsening disability (aOR = 2.38; 95% CI, 1.43–3.97), and presence of respiratory illness (aOR = 2.44; 95% CI, 1.67–3.56).
Conclusion
PCC were found in approximately one-fifth to one-third of participants with the presence of underlying medical conditions being a common risk factor in all outcomes. Advocacy for the prioritization of interventions such as vaccination of persons with underlying medical conditions could consequently result in a reduction in the risk of PCC.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12889-025-25863-7.
Keywords: Post-COVID conditions, Risk factors, Kenya
Background
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the pathogen responsible for the coronavirus disease 2019 (COVID-19) that emerged in late 2019, continues to spread with more than 767 million people reportedly infected globally as of 28th June 2023 [1]. The actual number of infected persons is likely to be higher due to inadequate testing especially in areas with weak surveillance and diagnostic systems, waning attention towards the virus, and in asymptomatic persons with a self-perception of low risk of infection. Serosurveys conducted to estimate the extent of exposure to SARS-CoV-2 suggest a higher level of exposure to this virus than PCR-based prevalence estimates [2–4]. COVID-19 vaccines that became available in early 2021 have been an effective tool against severe COVID-19 disease and death [5, 6] and controlling transmission spikes. In Kenya, as elsewhere, several waves occasioned by different variants were witnessed in the period 2020–2021 [7]. However, concerns about new virulent mutants of the virus that could potentially escape available vaccines persist, necessitating the need for heightened surveillance of the virus.
Although there have been advancements leading to a better understanding of the epidemiology of COVID-19 disease, including risk factors for severe disease and death [8–11], the impact of the disease on survivors is yet to be well described. While the majority of COVID-19 symptoms last an average of 1–3 weeks after onset [12, 13], a substantial proportion of patients have reported being unwell at post-discharge follow-ups done several months after infection [14, 15]. Recent reports suggest that 10–20% of COVID-19 disease survivors suffer persistent symptoms, or develop new ones, or are physically or psychologically impaired, or are unable to conduct activities to the same levels prior to infection, among other challenges [14–17]. This emerging phenomenon has been defined as Long COVID, or long-haulers or post-COVID conditions (PCC) [18–20].
Whereas there is considerable evidence that people with certain risk factors, such as high blood pressure, smoking, diabetes, and obesity, are more likely to have severe COVID-19 illness [8–10], there is limited congruent data on the association between these risk factors and PCC. As COVID-19 likely transitions to an endemic phase and the population affected by the disease grows, it is paramount to establish the burden and risk factors of PCC [14]. This is critical for the development of guidelines that can assist healthcare providers in the management of PCC.
Using an ongoing longitudinal population based infectious disease surveillance (PBIDS) platform established in 2006 in Kenya, we followed up two cohorts of participants — one in an urban, and the other in a rural setting — with previously laboratory confirmed SARS-CoV-2 infection to assess presence and risk factors of PCC.
Methods
The Kenya Medical Research Institute (KEMRI), in collaboration with the US Centers for Disease Control and Prevention (CDC), established PBIDS in Asembo, Siaya County in rural western Kenya, and in Kibera in Nairobi County. The two sites have differing characteristics, as described previously [21]. In brief, Asembo is a rural region in western Kenya whose residents are primarily subsistence farmers and fishermen. The area has a low population density of 325 persons per square kilometer. In contrast, Kibera is an informal settlement with a high population density of 77,000 persons per square kilometer, and inhabitants have limited access to clean, safe water, and no sewerage system. Residents work as casual laborers in the city or engage in small-scale businesses [22].
PBIDS collects data through active household surveillance involving regular home visits, and passive health facility surveillance of participants who receive free medical care for all acute, potentially infectious conditions. Each site has one centrally located health facility. The surveillance focuses on acute respiratory, acute febrile and diarrheal illnesses. As of 1 st July 2022, Asembo had 36,174 participants under follow-up and Kibera 23,521. All participants in the sites are assigned unique identification numbers that allow for their longitudinal tracking and linking of data.
Identification of COVID-19 cases
The first case of COVID-19 was reported in Kenya on 13th March 2020, and transmission continued in subsequent months in several waves [7]. From 1 st May 2020, PBIDS was leveraged to conduct surveillance of the SARS-CoV-2 infections among participants in Asembo and Kibera. All patients with severe illness (severe acute respiratory illness (see definitions in Supplementary Table 1)) presenting at the health facilities had nasopharyngeal/oropharyngeal (NP/OP) swabs collected and tested for SARS-CoV-2 using real-time reverse transcriptase-polymerase chain reaction (RT-PCR) at the CDC-supported KEMRI laboratories in Nairobi and Kisumu. The inclusion criteria for specimen collection and testing were expanded from 1 st September 2020 to include patients presenting with mild illnesses such as acute respiratory illness and acute febrile illness. Additionally, contacts of laboratory confirmed SARS-CoV-2 cases were invited to the health facilities for testing. Data on clinical presentations were systematically collected in real-time at the health facilities using standardized electronic data collection tools. SARS-CoV-2 results were linked to the patients’ respective clinical and demographic records and relayed back to individual patients as well as the Public Health Emergency Operations Centre at the Kenyan Ministry of Health (MoH). Test-positive cases were managed based on prevailing protocols set by the Kenyan MoH [23]. Kibera and Asembo detected their first SARS-CoV-2 cases on 6th May and 15th September 2020, respectively.
Follow-up of COVID-19 cases for PCC
Using the individual participant identification numbers, we selected all previously laboratory-confirmed SARS-CoV-2 positive cases from the PBIDS database for follow-up. The single follow-up was conducted between 13th October and 22nd November 2021. We adapted a survey tool known as the COVID-19 long term protocol from the International Severe Acute Respiratory and emerging Infection Consortium (ISARIC) [24] and programmed it onto tablets. For all selected cases willing to participate, trained field workers consented, enrolled and collected data through home visits. Data collected included the presence of symptoms during the follow-up visit, and the participant’s physical and mental health prior to, and after, the laboratory-confirmed SARS-CoV-2 infection. The surveys were administered in local languages used in the sites.
Data analysis
Participant’s baseline characteristics including education level, socioeconomic status and household size obtained from existing PBIDS databases were linked to the follow-up dataset using the unique participant’s identification numbers. For descriptive analysis, data were presented using medians, frequencies and percentages. We used logistic regression models adjusted for clustering at the household level to assess factors associated with four different outcomes identified as presence of respiratory symptoms, self-reported non-recovery following SARS-CoV-2 infection, psychological distress and worsening disability. Respiratory symptoms assessed during follow-up included presence of a persistent cough, chest pain, or shortness of breath. Self-reported non-recovery was measured by asking participants whether they felt that they had fully recovered from their SARS-CoV-2 infection. Psychological distress was defined as any experience of anxiety or depression following PCR-detection of SARS-CoV-2 infection. Worsening disability was defined as a new onset of, or an increase in difficulty in mobility, self-care, conducting usual activities, or discomfort since infection with SARS-CoV-2. PCC was defined as the presence of any of the four outcomes at least 1 month after SARS-CoV-2 infection. The predictors of the models were individuals’ age, sex, highest education level, socioeconomic status, household size, COVID-19 vaccination status (recommended by Kenyan MoH for persons ≥ 18 years at the time of the study [25]), site, and time (in months) since SARS-CoV-2 infection to follow up. Other predictors included the presence of underlying medical conditions (such as hypertension, diabetes, cancer, and cardiovascular disease) and severity of illness at the time of COVID-19 diagnosis classified as severe (SARI), mild (ARI and AFI) and asymptomatic persons (supplementary Table 1). All factors with a p-value ≤ 0.2 in the univariable logistic regression models were included in the multivariable models. Age and sex were considered potential confounders a priori in each model. Psychological distress and worsening disability required self-assessment for these outcomes; hence included persons aged ≥ 18 years only. Stata V17.0 software, (StataCorp, Texas US) was used in the analysis.
Ethical consideration
The PBIDS protocol and study instruments were reviewed and approved by the KEMRI scientific and ethics review unit (#2761) and received ethical reliance from the Washington State University institutional review board and US CDC IRB reliance approval (#6775). Written informed consent or assent was required of all participants for the collection of the follow-up survey data.
Results
Participant enrollment
After initiation of SARS-CoV-2 surveillance and detection of their first cases, both sites reported SARS-CoV-2 infections in all subsequent months with a monthly median case count of 21 (interquartile range (IQR) 7–25) in Asembo and 18 (IQR 13–33) in Kibera. SARS-CoV-2 PCR-positivity peaked in December 2020 – January 2021 and June – August 2020 for Asembo and Kibera, respectively (Supplementary Fig. 1). There were 1,053 participants (475 in Asembo and 578 in Kibera) who had PCR-confirmed SARS-CoV-2 infection between 1 st May 2020 to 30th September 2021. At follow-up, 8/475 (1.7%) and 3/578 (0.5%) from Asembo and Kibera, respectively, had died (Fig. 1). Of the survivors, 398 (85.2%) from 350 households in Asembo and 434 (75.5%) from 379 households in Kibera were found and enrolled in the follow-up survey. Reasons for non-participation in the two sites combined included those who had relocated or were away on travel (106/1,042, 10.2%), untraceable participants (57/1,042, 5.5%), and those away in school (40/1,042, 3.8%). Of note, the untraceable participants were more frequently from Kibera (54/575, 9.4%) than Asembo (3/467, 0.6%), (P-value < 0.001). There were no significant differences in the other aforementioned reasons for non-participation by site. Of the 839 total participants that were located, 7 (6 from Asembo and 1 from Kibera) declined to participate in the follow-up survey.
Fig. 1.
Flowchart for SARS-CoV-2 cases followed-up in Asembo and Kibera PBIDS sites in Kenya, October - November 2021
Demographic and clinical characteristics of survey participants
Of the 832 participants, 494 (59.3%) were female, 687 (82.6%) had at least primary level of education, and 5/511 (1.0%) were vaccinated (Table 1). Asembo had significantly more females in the survey than Kibera (63.1% vs. 56.0%, p-value = 0.038). The median age of the participants was 24 years (range 0–94). Asembo had a significantly higher proportion of individuals aged ≥ 50 years compared to Kibera (25.9% vs. 9.9%, p-value < 0.001). The following symptoms were more prevalent at COVID-19 diagnosis among the Asembo participants compared to those from Kibera; fever (53.0% vs. 24.0%), headache (72.4% vs. 45.6%), cough (71.4% vs. 59.0%) and muscle pain (36.9% vs. 15.2%). Chest pain was significantly more prevalent in Kibera (16.8%) than in Asembo patients (9.5%), (p-value = 0.002).
Table 1.
Demographic and clinical characteristics of COVID-19 follow up participants in Asembo and Kibera, Kenya, 2021
| Characteristic | Overall (N = 832) | Asembo (N = 398) | Kibera (N = 434) | P-value |
|---|---|---|---|---|
| n (%) | n (%) | n (%) | ||
| Demographics | ||||
| Age | ||||
| < 5 years | 107 (12.9) | 54 (13.6) | 53 (12.1) | 0.000 |
| 5–17 years | 214 (25.7) | 95 (23.9) | 119 (27.4) | |
| 18–29 years | 152 (18.3) | 59 (14.8) | 93 (21.4) | |
| 30–39 years | 122 (14.7) | 47 (11.8) | 75 (17.3) | |
| 40–49 years | 93 (11.2) | 40 (10.1) | 53 (12.2) | |
| 50–59 years | 73 (8.8) | 41 (10.3) | 32 (7.4) | |
| 60 + years | 71 (8.5) | 62 (15.6) | 9 (2.1) | |
| Gender: Female | 494 (59.3) | 251 (63.1) | 243 (56.0) | 0.038 |
| Household size 1–4 | 247 (29.7) | 118 (29.7) | 129 (29.7) | 0.933 |
| 5–6 | 292 (35.1) | 142 (35.7) | 150 (34.6) | |
| 7+ | 293 (35.2) | 138 (34.7) | 155 (35.7) | |
|
Time since diagnosis to follow up < 6 months 6 + months |
453 (54.5) | 253 (63.6) | 200 (46.1) | < 0.001 |
| 379 (45.5) | 145 (36.4) | 234 (53.9) | ||
| Education | ||||
| Not of school going age | 69 (8.3) | 37 (9.3) | 32 (7.4) | |
| Not completed formal education or training | 72 (8.7) | 58 (14.6) | 14 (3.2) | < 0.001 |
| Primary education | 425 (51.1) | 193 (48.5) | 232 (53.6) | |
| Secondary/High school | 197 (23.7) | 80 (20.1) | 117 (27.0) | |
| Post-Secondary | 65 (7.8) | 28 (7.0) | 37 (8.6) | |
| Prefer not to say | 4 (0.5) | 2 (0.5) | 2 (0.5) | |
| Covid-19 Vaccinated a | 5/511 (1.0) | 2/249 (0.8) | 3/262 (1.2) | 1.001 |
| Underlying medical conditions | 182 (21.9) | 94 (23.6) | 88 (20.3) | 0.282 |
| Illness severity at diagnosis b | ||||
| No symptoms | 106 (12.7) | 40 (10.0) | 66 (15.2) | 0.187 |
| Mild | 695 (83.5) | 341 (85.7)) | 354 (81.6 | |
| Severe | 31 (3.7) | 17 (4.3) | 14 (3.2) | |
| Symptoms at COVID-19 diagnosis | ||||
| Fever | 315 (37.9) | 211 (53.0) | 104 (24.0) | < 0.001 |
| Headache | 486 (58.4) | 288 (72.4) | 198 (45.6) | < 0.001 |
| Cough | 540 (64.9) | 284 (71.4) | 256 (59.0) | < 0.001 |
| Difficulty Breathing | 30 (3.6) | 15 (3.8) | 15 (3.5) | 0.854 |
| Chest pain | 111 (13.3) | 38 (9.5) | 73 (16.8) | 0.002 |
| Vomiting | 21 (2.5) | 11 (2.8) | 10 (2.3) | 0.826 |
| Diarrhea | 15 (1.8) | 5 (1.3) | 10 (2.3) | 0.304 |
| Muscle pain | 212 (25.5) | 146 (36.9) | 66 (15.2) | < 0.001 |
| Joint pain | 210 (25.2) | 97 (24.4) | 113 (26.0) | 0.632 |
aVaccine recommended for persons > = 18 years at time of study
bRefer to Supplementary Table 1 for definitions
Among those with the PCC outcomes, 63.2–75.0% were females, and 85.5–88.4% had mild illnesses at the time of their COVID-19 diagnosis (Supplementary Table 2).
The time interval since SARS-CoV-2 infection to follow-up varied for the sites (Fig. 2). For Asembo the time interval ranged from 1 to 13 months (median = 5 months; Interquartile range, IQR 4–8) while for Kibera the range was 1–18 months (median = 7 months; IQR 3–12).
Fig. 2.
Distribution of participants' interval time in months from SARS-CoV-2 diagnosis date to follow- up date in Asembo and Kibera, Kenya, 2021
Prevalence of psychological distress, self-reported non-recovery to COVID-19 illness, presence of respiratory symptoms and worsening disability
The overall prevalence of psychological distress was 29.7% (152/511), for worsening disability was 21.9% (112/511), for reported non-recovery from COVID-19 was 20.8% (165/793), while for respiratory symptoms at follow up was 20.9% (174/832) (Table 2). Among persons < 18 years, 34.0% (109/321) had at least one PCC outcome, while corresponding prevalence for persons ≥ 18 years was 56.9% (291/511). Prevalence of psychological distress increased with age, ranging from 16.4% in persons 18–29 years to a high of 46.5% among persons aged ≥ 60 years. Psychological distress was also more prevalent in persons followed-up within 6 months of infection compared to those with longer periods (35.7% vs. 23.2%, p = 0.002). Prevalence of all outcomes was more common in females than males; significantly more so in those with a worsening disability (25.4% vs. 15.5%, p < 0.05). The prevalence of PCC ranged from 6.3% − 13.3% among participants with underlying medical conditions.
Table 2.
Prevalence of psychological distress, worsening disability, reported non-recovery from COVID-19 and persistence of respiratory symptoms by participants’ characteristics in Asembo and Kibera Kenya, 2021
| Psychological distress | Worsening disability | Reported non-recovery* | Presence of resp. illness | |||||
|---|---|---|---|---|---|---|---|---|
| n/N | % | n/N | % | n/N | % | n/N | % | |
| Age | ||||||||
| <5 | - | - | 30/106 | 28.3 | 26/107 | 24.3 | ||
| 5–17 | - | - | 40/206 | 19.4 | 35/214 | 16.4 | ||
| 18–29 | 25/152 | 16.4 | 24/152 | 15.8 | 22/145 | 15.2 | 29/152 | 19.1 |
| 30–39 | 34/122 | 27.9 | 28/122 | 23.0 | 19/115 | 16.5 | 26/122 | 21.3 |
| 40–49 | 32/93 | 34.4 | 21/93 | 22.6 | 17/86 | 19.8 | 21/93 | 22.6 |
| 50–59 | 28/73 | 38.4 | 18/73 | 24.7 | 14/68 | 20.6 | 11/73 | 15.1 |
| 60+ | 33/71 | 46.5 | 21/71 | 29.6 | 23/67 | 34.3 | 26/71 | 36.6 |
| Gender | ||||||||
| Male | 45/181 | 24.9 | 28/181 | 15.5 | 57/323 | 17.6 | 64/338 | 18.9 |
| Female | 107/331 | 32.3 | 84/331 | 25.4 | 108/470 | 23.0 | 110/494 | 22.3 |
| Time since COVID-19 dx | ||||||||
| <6 months | 95/266 | 35.7 | 72/266 | 27.1 | 97/438 | 22.1 | 96/453 | 21.2 |
| >=6 months | 57/246 | 23.2 | 40/246 | 16.3 | 68/355 | 19.2 | 78/379 | 20.6 |
| Household size | ||||||||
| 1–4 | 37/161 | 23.0 | 32/161 | 19.9 | 56/235 | 23.8 | 47/247 | 19.0 |
| 5–6 | 58/179 | 32.4 | 36/179 | 20.1 | 67/280 | 23.9 | 71/292 | 24.3 |
| >=7 | 57/172 | 33.1 | 44/172 | 25.6 | 42/278 | 15.1 | 56/293 | 19.2 |
| Illness severity at diagnosis | ||||||||
| No symptoms | 13/62 | 21.0 | 6/62 | 9.7 | 15/83 | 18.1 | 13/106 | 12.1 |
| Mild | 130/426 | 30.5 | 99/426 | 23.2 | 144/679 | 21.2 | 153/695 | 22.0 |
| Severe | 9/24 | 37.5 | 7/24 | 29.2. | 6/31 | 19.4 | 8/31 | 25.8 |
| Underlying medical conditions | 68/511 | 13.3 | 55/511 | 10.8 | 50/793 | 6.3 | 53/832 | 6.4 |
| COVID-19 vaccinated** | 0/511 | 0.0 | 1/511 (0.2) | 0.2 | 1/481 | 0.2 | 1/511 | 0.2 |
| Overall | 152/511 | 29.7 | 112/511 | 21.9 | 165/793 | 20.8 | 174/832 | 20.9 |
*Those reporting “Don’t know” are excluded from denominator
**COVID-19 vaccination was only recommended for persons ≥18 years at the time of the study
Factors associated with psychological distress, self-reported non-recovery to COVID-19 illness, presence of respiratory symptoms and worsening disability
Factors with a p value < 0.2 in the univariate analysis (Supplementary Table 3) were included in the multivariable models. In the four multivariable models, the presence of underlying medical conditions was significantly associated with all outcomes; for psychological distress the adjusted Odds Ratio (aOR) was 1.87 (95% CI, 1.17–2.99), for worsening disability aOR, 2.58; (95% CI, 1.54–4.34), for reported non-recovery aOR, 1.93; (95% CI, 1.24–3.02) and for presence of respiratory symptoms aOR, 1.82; (95% CI, 1.16–2.87), (Table 3). Household sizes with > 4 persons were associated with increased reports of psychological distress; aOR, 1.97, (95% CI, 1.13–3.44) for households with 5–6 persons, and aOR, 1.89, (95% CI, 1.10–3.22) for those with > 6 persons. Living in Asembo was associated with increased odds of psychological distress (aOR, 2.23; 95% CI, 1.42–3.53), worsening disability (aOR, 2.38; 95%CI, 1.43–3.97) and persistence of respiratory symptoms (aOR, 2.44; 95% CI, 1.67–3.56). Females had significantly higher odds than males of reporting not having fully recovered after SARS-CoV-2 infection (aOR, 1.47; 95% CI, 1.01–2.13). A longer time interval since SARS-CoV-2 infection, severity of illness at the time of COVID-19 diagnosis, and education level attained were not associated with any of the four outcomes assessed.
Table 3.
Risk factors associated with psychological distress, worsening disability, reported non-recovery and presence of respiratory illnesses among persons infected with SARS-CoV-2 in population-based infectious disease surveillance in Asembo and Kibera, Kenya, 2021
| Characteristic | Psychological distress **OR (95% CI) |
Worsening disability **OR (95% CI) |
Reported non-recovery **OR (95% CI) |
Presence of respiratory illness* **OR (95% CI) |
|
|---|---|---|---|---|---|
| Age Group | < 5yrs | - | - | Ref | Ref |
| 5-17yrs | - | - | 0.81 (0.39–1.70) | 1.04 (0.45–2.39) | |
| 18-29yrs | Ref | Ref | 0.60 (0.25–1.46) | 1.35 (0.53–3.44) | |
| 30-39yrs | 1.62 (0.84–3.10) | 1.64 (0.82–3.31) | 0.52 (0.22–1.26) | 1.43 (0.58–3.57) | |
| 40-49yrs | 1.92 (1.00–3.71.00.71) | 1.31 (0.64–2.68) | 0.66 (0.27–1.62) | 1.32 (0.50–3.48) | |
| 50-59yrs | 2.27 (1.11–4.60) | 1.29 (0.58–2.85) | 0.69 (0.27–1.80) | 0.70 (0.24–2.03) | |
| 60+ | 2.13 (0.94–4.79) | 1.02 (0.42–2.47) | 1.29 (0.54–3.06) | 1.52 (0.55–4.18) | |
| Gender | Male | Ref | Ref | Ref | Ref |
| Female | 1.25 (0.79–1.96) | 1.50 (0.88–2.56) | 1.47 (1.01–2.13) | 1.06 (0.73–1.54) | |
| Site | Kibera | Ref | Ref | - | Ref |
| Asembo | 2.23 (1.42–3.53) | 2.38 (1.43–3.97) | - | 2.44 (1.67–3.56) | |
| Household sizes (no. of people) | 1–4 | Ref | - | Ref | - |
| 5–6 | 1.97 (1.13–3.44) | - | 1.10 (0.71–1.72) | - | |
| 7+ | 1.89 (1.10–3.22) | - | 0.60 (0.37 −1.00) | - | |
| Time since SARS-CoV-2 diagnosis (Months) | <6 | Ref | Ref | - | - |
| 6+ | 0.65 (0.41–1.02) | 0.70 (0.42–1.16) | - | - | |
| Education | Not completed formal education | Ref | - | Ref | Ref |
| Pre-primary | - | - | 2.52 (0.94–6.75) | 1.93 (0.71–5.23) | |
| Primary education | 1.57 (0.67–3.71) | - | 1.73 (0.82–3.67) | 0.93 (0.50–1.71) | |
| Secondary/High school | 0.87 (0.34–2.23) | - | 1.49 (0.62–3.58) | 0.87 (0.43–1.75) | |
| Post-Secondary | 0.78 (0.26–2.34) | - | 0.93 (0.30–2.88) | 0.70 (0.28–1.73) | |
| Underlying medical condition | No | Ref | Ref | Ref | Ref |
| Yes | 1.87 (1.17–2.99) | 2.58 (1.54–4.34) | 1.93 (1.24–3.02) | 1.82 (1.16–2.87) | |
| Illness Severity at diagnosis | No symptoms | - | Ref | - | Ref |
| Mild symptoms | - | 1.74 (0.66–4.59) | - | 1.65 (0.83–3.30) | |
| Severe | - | 2.41 (0.65–8.96) | - | 1.92 (0.64–5.74) | |
| - | - | - | - | ||
*Defined as presence of persistent cough or difficult breathing or chest pain
**Adjusted OR accounting for household clustering
Discussion
Our findings document the existence of persisting or emerging physical and health conditions affecting nearly one-fifth to one-third of persons previously infected with SARS-CoV-2 in urban and rural Kenya. We combined data obtained from health facilities, laboratories, and households resulting in a rich dataset that enabled us to examine PCC and the associated risk factors in well-defined catchment settings in urban and rural Kenya. The participants’ follow-ups conducted after different months (ranging from 1 to 18 months) following infection with SARS-CoV-2 revealed that a substantial proportion of those previously infected continued to experience a variety of physical or psychological challenges. The four PCC outcomes that we evaluated included psychological distress, worsening disability, self-assessed non-recovery and presence of respiratory symptoms, with prevalence ranging from 20.8 to 29.7%. Among the risk factors examined, female gender, large household (with > 4 persons), and presence of underlying medical conditions were noted to be associated with increased odds of having PCC.
Several studies have demonstrated the role of underlying medical conditions such as diabetes and cardiovascular disease with regard to severe outcomes from acute COVID-19 disease; the risk of hospitalization and death is much higher in patients with such comorbidities in comparison with those without [26, 27]. For the case of PCC, we observed a significant association between the presence of underlying medical conditions and all four PCC outcomes assessed. Persons with these comorbidities were nearly twice as likely to suffer PCC compared to those without. These findings point to the prioritization of persons with underlying medical conditions for preventive measures such as vaccination to reduce the risk of SARS-CoV-2 infections, and subsequently, PCC.
We established that living in Asembo, which is a rural site, was associated with PCC in three of the four outcomes with participants at the site having twice the odds of experiencing PCC than the urban residents. Most of rural Kenya is characterized by poor health infrastructure. Although free healthcare was provided to PBIDS participants with potentially infectious illnesses at the study clinics, we have previously documented that numerous other considerations such as distance and transport costs affect access to healthcare [28]. Furthermore, scarcity of inpatient units and intensive care including oxygen supplementation and ventilators for those critically ill – even if they were able to afford it – is a more common phenomenon in rural than urban areas in Kenya. Patients in rural areas in need of advanced care often transfer to higher, advanced levels of care in major towns since this critical care is not available in rural Kenya [29]. However, these advanced health facilities are limited in number, and difficult to access for most rural poor participants. As the COVID-19 pandemic raged on in the second and third quarters of the year 2021, the health system in Kenya was stretched with hospital beds extremely difficult to find [30]. Despite efforts by county governments to upgrade their health infrastructure, critically ill patients, especially in rural parts of Kenya stayed at home due to the absence of in-patient services, or exorbitant hospitalization costs. This may explain why the site of residence was an important factor for PCC in our study.
We also observed that persons living in households with more than four people had increased odds of psychological distress. Potential difficulties to “social distance” or isolate may have resulted in concerns about health; the implication being that the large numbers of persons in the household would likely result in increased transmission, including re-infections within the households. This may explain the differences seen in smaller household sizes (≤ 4 persons) in comparison with larger ones.
It has been reported in studies from Korea and Moscow [15, 31] that as people age health concerns increase. In our study, the age effect was notable only in persons aged (50–59 years) who had increased odds for psychological distress. We did not find any differences in the other age groups and neither did we see any significant differences by age in the other outcomes. Consistent with some studies that have found differentials in PCC based on sex with females having greater risk than men [19], in our follow-up study, we saw sex differences in PCC for those with reported non-recovery from COVID-19 illness.
Some studies have shown mixed PCC effects irrespective of the degree of severity of illness at the time of diagnosis. Examples from the USA and Iran showed that hospitalized COVID-19 patients had a higher risk of PCC than those screened at outpatient settings or hospitalized for a short duration [18, 32]. Another study from Kerala in India found that asymptomatic patients or those with mild symptoms only, also suffered PCC during follow-up [33]. In this study, we included three categories of persons with differing levels of illness severity; first were those who had been in contact with a confirmed case even if they were asymptomatic, the second group were those that had only mild illnesses such as a fever or cough only, and lastly were those with severe acute respiratory illness. However, we did not see any difference in PCC outcomes based on the severity of illness of the patients at the time of diagnosis, though this might have been affected by small sample sizes in the sub-groups. It is possible that some of the asymptomatic persons or mild cases progressed into severe states or did not have any further effects after their infection. Since we did not closely monitor these cases through frequent home visits or in their isolation facilities, it is difficult to tell the patients’ sequelae.
At the two PBIDS sites, the time interval taken from diagnosis to follow-up varied. While it took up to 18 months in Kibera, for Asembo the longest interval was 13 months. The differences in the intervals were primarily occasioned by the periods at which SARS-CoV-2 detection first occurred at the two sites, with Kibera reporting cases much earlier than Asembo. The first cases were detected in May and September of 2020 in Kibera and Asembo respectively. At the onset of the pandemic in Kenya, Nairobi County in which Kibera is located, witnessed a surge in SARS-CoV-2 infections. This prompted the Kenyan government to institute tough restrictions including limiting movement in and out of Nairobi County [34]. However, over time, some of the restrictions were lifted, and travel across counties resumed and this likely led to increased transmission in other counties. Given that we conducted the survey simultaneously across the two sites, some participants in Kibera would inevitably have had a longer duration since infection to the follow up survey period.
Of the participants identified for follow-up, some could not be traced for an interview despite several attempts by field staff to locate them. This was notably more common in Kibera affecting nearly one-tenth of participants identified with SARS-CoV-2 compared with Asembo that had less than 1%. There are several possible explanations for this large discrepancy across the two sites. First PBIDS field workers reported that there was displacement of hundreds of families and demolition of their dwellings to create room for the construction of roads in the Kibera area. Several affected families may have moved out of the study area in search of new accommodation. Secondly, several individuals in Kibera are known to be solitary dwellers. These include people who moved to the city to work as casuals or temporary staff in the city’s industrial area. With the impact of the pandemic, many likely lost their jobs and relocated back to their home counties. Inf the absence of any family members to provide information on such participants, it was impossible to locate them. Lastly, Asembo is a rural area with many participants living in family or ancestral land. Locating participants absent from their homes during interview visits is much easier than for urban residents.
This study had some limitations. First, we conducted only one single follow-up per participant to document the possible presence of PCC. Additional follow-ups might potentially have yielded additional data showing how participants progressed after the first contact. However, we were limited in resources to conduct additional visits. Secondly, it is possible that some of the outcomes, such as psychological distress may have been driven by other unmeasured factors, such as loss of livelihoods due to the pandemic, other comorbidities, or movement restriction measures applied by the Kenyan government, as opposed to individuals’ SARS-CoV-2 infection. We saw an age effect for persons aged 50–59 years reporting psychological distress, but this was not observed for any other age group. We couldn’t completely rule out residual confounding. Follow up data from a comparison group of non-COVID-19 patients would have been required for this assessment to unequivocally associate the outcomes to SARS-Cov-2 infection. Thirdly, there exists a possibility of recall bias for symptoms experienced given that these were self-reported. We initiated follow-ups early (from 1 month) to mitigate the effects of recall bias. This timing may not have adhered to the recommended 3-month lag. Lastly, given that possible PCC symptoms are numerous, this study could not cover all. Some commonly reported symptoms of Long COVID (in the US) are fatigue and brain fog, yet these were not measured directly in our assessment.
Conclusions
Although the PCC outcomes assessed in this study are not a comprehensive list of all possible PCC symptoms, our data provides rich evidence on the presence of PCC across two communities in Kenya, underpinning the critical need for appropriate guidelines to address this problem among COVID-19 survivors.
Supplementary Information
Acknowledgements
We would like to thank the PBIDS participants in Kibera and Asembo sites for their continued participation in the surveillance platform. Secondly, we would like to thank the field teams that conducted home visits to obtain data, as well as health facility surveillance teams for helping identify COVID-19 cases and contacts.
Disclaimer
The findings and conclusions in this report are those of the authors and do not necessarily represent the official views of the U.S. Centers for Disease Control and Prevention.
Abbreviations
- CDC
US Centers for Disease Control and Prevention
- CI
Confidence interval
- IQR
Interquartile range
- KEMRI
Kenya Medical Research Institute
- MoH
Ministry of Health
- NP/OP
Nasopharyngeal/Oropharyngeal swabs
- PBIDS
Population based infectious disease surveillance
- PCC
Post–COVID Conditions
- RT-CPR
Real–time reverse Transcriptase–Polymerase Chain Reaction
- SARS-CoV-2
Severe Acute Respiratory Syndrome Coronavirus 2
Authors’ contributions
GB, PKM, TL, conceptualized the project, developed protocol and data collection tools, supervised data collection, analyzed and wrote manuscript. AA, BO, GA, CO developed data collection tools, analyzed data, reviewed manuscript. AO, DO, TK, GOA supervised data collection, participated in data review and analysis, and reviewed the manuscript. TL, AH, PM, CN reviewed the data collection tools, participated in the data analysis and reviewed manuscript.
Funding
This study was funded by the U.S. Centers for Disease Control and Prevention through Cooperative Agreement #6U01GH002143, and the Bill and Melinda Gates Foundation (Grant #INV-021779).
Data availability
Although no personal identifiers are included in this manuscript, the datasets generated during the current study contain these identifiers and are therefore not publicly available. However, a de-identified dataset may be availed by the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
The study protocol and data collection instruments were reviewed and approved by the KEMRI scientific and ethics review unit (#2761) and received ethical reliance from the Washington State University institutional review board and US CDC IRB reliance approval (#6775). Written informed consent or assent was required of all participants for the collection of the follow-up survey data and linking to other PBIDS databases.
Consent for publication
No personal identifiers are included in the manuscript. All coauthors have provided consent for the publication of this manuscript.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.World Health Organization. WHO Coronavirus (COVID-19) Dashboard. Available at https://www.who.int/news-room/questions-and-answers/item/coronavirus-disease-(covid-19)-post-covid-19-condition.
- 2.Bendavid E, et al. COVID-19 antibody Seroprevalence in Santa Clara County, California. Int J Epidemiol. 2021;50(2):410–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Muller SA, et al. Seroprevalence and risk factors of COVID-19 in healthcare workers from 11 African countries: a scoping review and appraisal of existing evidence. Health Policy Plan. 2022;37(4):505–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Munywoki PK, Nasimiyu C, Alando MD, Otieno N, Ombok C, Njoroge R, Kikwai G, Odhiambo D, Osita MP, Ouma A, Odour C, Juma B, Ochieng CA, Mutisya I, Ngere I, Dawa J, Osoro E, Njenga MK, Bigogo G, Munyua P, Lo TQ, Hunsperger E, Herman-Roloff A, et al. Seroprevalence and risk factors of SARS-CoV-2 infection in an urban informal settlement in Nairobi, Kenya, 2020. F1000Res. 2021;10:853. 10.12688/f1000research.72914.2. PMID: 35528961 PMCID: PMC9065925. [DOI] [PMC free article] [PubMed]
- 5.Watson OJ, et al. Global impact of the first year of COVID-19 vaccination: a mathematical modelling study. Lancet Infect Dis. 2022;22(9):1293–302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Baker TB, et al. The relationship of COVID-19 vaccination with mortality among 86,732 hospitalized patients: Subpopulations, patient Factors, and changes over time. J Gen Intern Med. 2023;38(5):1248–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Nasimiyu C, Matoke-Muhia D, Rono GK, Osoro E, Obado DO, Mwangi JM, Mwikwabe N, Thiong'o K, Dawa J, Ngere I, Gachohi J, Kariuki S, Amukoye E, Mureithi M, Ngere P, Amoth P, Were I, Makayotto L, Nene V, Abworo EO, Njenga MK, Seifert SN, Oyola SO, et al. Imported SARS-COV-2 variants of concern drove spread of infections across kenya during the second year of the pandemic. medRxiv [Preprint]. 2022:22271467. 10.1101/2022.02.28.22271467. Update in: This article has been published with 10.3390/covid2050044. PMID: 35262086; PMCID: PMC8902869.
- 8.Bertagnolio S, et al. Clinical features of, and risk factors for, severe or fatal COVID-19 among people living with HIV admitted to hospital: analysis of data from the WHO global clinical platform of COVID-19. Lancet HIV. 2022;9(7):e486–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Booth A, et al. Population risk factors for severe disease and mortality in COVID-19: A global systematic review and meta-analysis. PLoS ONE. 2021;16(3):e0247461. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Choi JH, Choi SH, Yun KW. Risk factors for severe COVID-19 in children: A systematic review and Meta-Analysis. J Korean Med Sci. 2022;37(5):e35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wingert A, et al. Risk factors for severity of COVID-19: a rapid review to inform vaccine prioritisation in Canada. BMJ Open. 2021;11(5):e044684. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lechien JR, et al. Clinical and epidemiological characteristics of 1420 European patients with mild-to-moderate coronavirus disease 2019. J Intern Med. 2020;288(3):335–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tenforde MW, et al. Symptom duration and risk factors for delayed return to usual health among outpatients with COVID-19 in a multistate health care systems Network - United States, March-June 2020. MMWR Morb Mortal Wkly Rep. 2020;69(30):993–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Nalbandian A, et al. Post-acute COVID-19 syndrome. Nat Med. 2021;27(4):601–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kim Y, et al. Post-acute COVID-19 syndrome in patients after 12 months from COVID-19 infection in Korea. BMC Infect Dis. 2022;22(1):93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chippa V, Aleem A, Anjum F. Postacute Coronavirus (COVID-19) Syndrome. [Updated 2024 Mar 19]. In: StatPearls. Treasure Island (FL): StatPearls Publishing; 2025. Available from: https://www.ncbi.nlm.nih.gov/books/NBK570608/. [PubMed]
- 17.Mahmud R, et al. Post-COVID-19 syndrome among symptomatic COVID-19 patients: A prospective cohort study in a tertiary care center of Bangladesh. PLoS ONE. 2021;16(4):e0249644. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Asadi-Pooya AA, et al. Risk factors associated with long COVID syndrome: A retrospective study. Iran J Med Sci. 2021;46(6):428–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Sylvester SV, Rusu R, Chan B, Bellows M, O'Keefe C, Nicholson S. Sex differences in sequelae from COVID-19 infection and in long COVID syndrome: a review. Curr Med Res Opin. 2022;38(8):1391-399. https://doi.org/10.1080/03007995.2022.2081454. Epub 2022 Jun 20. PMID: 35726132. [DOI] [PubMed]
- 20.World Health Organization. G., Coronavirus disease (COVID-19): Post COVID-19 condition. Available at https://www.who.int/news-room/questions-and-answers/item/coronavirus-disease-(covid-19)-post-covid-19-condition.
- 21.Feikin DR, et al. The burden of common infectious disease syndromes at the clinic and household level from population-based surveillance in rural and urban Kenya. PLoS ONE. 2011;6(1):e16085. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Odhiambo FO, et al. Profile: the KEMRI/CDC health and demographic surveillance System–Western Kenya. Int J Epidemiol. 2012;41(4):977–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ministry of Health, K., Interim guidelines on management of COVID-19 in Kenya. 2020. Available at https://www.health.go.ke/wp-content/uploads/2020/06/Updated-Case-Management-Guidelines-26_03_20-1.pdf
- 24.ISARIC. COVID-19 long term protocol. 2020. Available at https://isaric.org/wp-content/uploads/2020/12/Tier-1-Initial-Follow_up_survey.pdf.
- 25.Ministry of Health, K., Frequently asked questions on COVID-19 vaccination in Kenya. 2021. Available at https://www.health.go.ke/wp-content/uploads/2021/08/Frequently-asked-questions-on-COVID-19-Vaccination.pdf.
- 26.Bepouka B, et al. Diabetes mellitus is associated with higher COVID-19 mortality rates in Sub-Saharan africa: A systematic review and Meta-analysis. Cureus. 2022;14(7):e26877. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Sahni S, et al. Impact of metabolic and cardiovascular disease on COVID-19 mortality: A systematic review and meta-analysis. Diabetes Metab Syndr. 2021;15(6):102308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bigogo G, et al. Health-seeking patterns among participants of population-based morbidity surveillance in rural Western kenya: implications for calculating disease rates. Int J Infect Dis. 2010;14(11):e967–73. [DOI] [PubMed] [Google Scholar]
- 29.Ministry of Health, K., Kenya Health sector Referral Strategy (2014–2018). 2013. Available at https://www.measureevaluation.org/pima/referral-systems/referral-strategy.
- 30.Barasa EW, Ouma PO, Okiro EA. Assessing the hospital surge capacity of the Kenyan health system in the face of the COVID-19 pandemic. PLoS ONE. 2020;15(7):e0236308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Pazukhina E, et al. Prevalence and risk factors of post-COVID-19 condition in adults and children at 6 and 12 months after hospital discharge: a prospective, cohort study in Moscow (StopCOVID). BMC Med. 2022;20(1):244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Yoo SM, et al. Factors associated with Post-Acute sequelae of SARS-CoV-2 (PASC) after diagnosis of symptomatic COVID-19 in the inpatient and outpatient setting in a diverse cohort. J Gen Intern Med. 2022;37(8):1988–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Anjana NKN, et al. Manifestations and risk factors of post COVID syndrome among COVID-19 patients presented with minimal symptoms - A study from Kerala, India. J Family Med Prim Care. 2021;10(11):4023–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Kenya Law. Public health (Covid-19 restriction of movement of persons and related measures) rules. Legal notice 50 of 2020. Available at https://new.kenyalaw.org/akn/ke/act/ln/2020/50/eng@2022-12-31.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Although no personal identifiers are included in this manuscript, the datasets generated during the current study contain these identifiers and are therefore not publicly available. However, a de-identified dataset may be availed by the corresponding author on reasonable request.