Respiratory tract infections at the Grand Magal de Touba: a seven-year study

Abstract

The Grand Magal of Touba (GMT) is the largest religious gathering in West Africa, and, like any large mass gathering, presents a risk of transmission of infectious diseases, mainly respiratory tract infections. Respiratory pathogen carriage was assessed by qPCR between 2017 and 2023 in healthy and ill pilgrims. The study included 1067 pilgrims comprizing 328 healthy controls, 298 pilgrims with respiratory symptoms who did not consult and 441 patients who consulted for respiratory symptoms. Among healthy controls, 7% tested positive for rhinovirus, 4% for endemic coronaviruses, 45% for Haemophilus spp. and 29% for Staphylococcus aureus. Among non-consulting ill pilgrims, 14% tested positive for rhinovirus, 8% for endemic coronaviruses, 45% for Haemophilus spp., 31% for S. aureus and 32% for Streptococcus pneumoniae. Among consulting patients, 29% tested positive for influenza A, 14% for respiratory syncytial virus, 9% for rhinovirus, 62% for Haemophilus spp., 47% for S. pneumoniae and 39% for Moraxella catarrhalis. Co-infection with viruses and bacteria was 12% in healthy controls, 23% in non-consulting ill pilgrims and 51% in consulting patients. Female gender was independently associated with a higher risk for respiratory symptoms. Symptomatic individuals were 53 times more likely to test positive for influenza A virus, 2 times for S. pneumoniae and M. catarrhalis and 3 times more likely to have a bacteria-virus co-infection. When comparing non-consulting ill pilgrims and consulting patients, the latter were 29 times more likely to test positive for influenza B virus, 16 times for respiratory syncytial virus, 14 times for influenza A virus and 3 times for M. catarrhalis. Influenza virus, respiratory syncytial virus, S. pneumoniae and M. catarrhalis could play a major role in the pathogenesis of respiratory infections at the GMT.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-16489-1.

Introduction

Every 18th day of the month of Safar (Muslim calendar), in Senegal, the Grand Magal of Touba (GMT) is celebrated, commemorating the departure into exile of Sheikh Ahmadou Bamba, the founder of Mouridism, in 1895, in Gabon. It takes place in the holy city of Touba (Diourbel department) and attracts some 4–5 million people from all over Senegal and beyond. The event is the largest religious gathering in West Africa, and, like any large mass gathering, presents a risk of transmission of infectious diseases, mainly respiratory tract infections¹. Prospective cohort studies conducted on 535 pilgrims from Ndiop and Dielmo villages (2017–2021) showed that 55% had respiratory symptoms, with the main pathogens acquired following participation in the GMT being: Haemophilus spp. (18.9%), Streptococcus pneumoniae (14.1%), Klebsiella pneumoniae (13%), Staphylococcus aureus (12%), rhinovirus (10%), endemic coronaviruses (6%) and Moraxella catarrhalis (5%)². Similarly, high prevalence of respiratory pathogens was also observed among pilgrims consulting for respiratory symptoms at the Mbacké healthcare center, close to Touba (2018–2021), with Haemophilus spp. (73%), S. pneumoniae (51%), M. catarrhalis (46%), S. aureus (22%), influenza A virus (26%), respiratory syncytial virus (16%), rhinovirus (8%) and influenza B virus (7%) the most frequent pathogens isolated from respiratory samples³. In this study, we extended our investigation until 2023 with the aim to investigate the relationship between the presence of respiratory pathogens and the occurrence of respiratory symptoms.

Materials and methods

Study population

The study involved cohorts of pilgrims departing from the villages of Dielmo (13° 43′ 22.07″ N, 16° 24′ 40.09″ W) and Ndiop (13° 41′ 08.01″ N, 16° 23′ 01. 01″ W) located in southern Senegal, in the Fatick region, and patients consulting during day time, at the Mbacké healthcare center (14° 47′ 56″ N, 15° 54′ 36″ W, 10 km from Touba). The Mbacké healthcare center was chosen because of its high rate of attendance during the 5 days of free GMT medical coverage, (1115 patients during the 2015 GMT, unpublished data). Pilgrims were identified by the nurses in charge of the primary healthcare centers in the two villages or by the medical staff at the Mbacké healthcare center and participated in the study on a voluntary basis. All pilgrims corresponding to these inclusion criteria were proposed to be enrolled in the study. All pilgrims enrolled in this study participated to the religious event. Participants (or their parents if they were minors) were asked to sign a written consent form.

Data collection

At the time of inclusion of cohort pilgrims (2017–2023), the procedure was to collect data using a standardized questionnaire 6 to 9 days after they returned to their village from Touba. Informed consent for testing was given and for children informed consent was obtained from a parent or legal guardian. The survey covered demographic data (age and sex), vaccination status, and respiratory symptoms during and after the pilgrimage, antibiotic consumption during the trip and hospitalization.

During each 5-day GMT free medical coverage period (2018–2023), the medical team at the Mbacké healthcare center completed a demographic and clinical questionnaire, including patients whose main complaint was respiratory symptoms. Three groups of individuals were retrospectively identified: (i) healthy controls: cohort pilgrims who did not report respiratory symptoms on returning to Touba; (ii) non-consulting ill pilgrims: cohort pilgrims who reported respiratory symptoms on returning to Touba but did not consult a healthcare practitioner for that reason; (iii): consulting patients: pilgrims who consulted a healthcare practitioner for respiratory symptoms. A participant was considered as “symptomatic” when he had at least one respiratory symptom (cough, rhinitis, dyspnea, sore throat). A fever was not required.

Sample collection

Cohort pilgrims were sampled 6 to 9 days following their return home and Mbacké patients were sampled when consulting. Samples were taken using commercial rigid cotton swab applicators (Medical Wire & Equipment, Wiltshire, UK, MW176S) inserted into the anterior nostrils or oropharynx, then placed in viral transport medium (Sigma Virocult^®). Samples were kept and transported at + 4 °C to the Dakar laboratory for storage in a − 80 °C freezer, then transferred to Marseille on dry ice for processing.

Identification of respiratory pathogens by PCR

To search for a panel of respiratory pathogens consisting of influenza A, influenza B, human rhinovirus, respiratory syncytial virus, human metapneumovirus, endemic human coronaviruses, SARS-CoV-2, adenovirus, S. pneumoniae, S. aureus, Haemophilus spp., K. pneumoniae, Bordetella pertussis, Mycoplasma pneumoniae and M. catarrhalis, we used in-house real-time PCR assay. Primers used are presented in Supplementary Table 1.

performed nucleic acid extraction (DNA and RNA) using the EZ1 Advanced XL (Qiagen, Hilden, Germany) with the Mini Virus Kit v2.0 (Qiagen) in accordance with the manufacturer’s recommendations. Genes were amplified using the LightCycler 480 Probes Master Kit (Roche Diagnostics, France) for bacteria and the Multiplex RNA Virus Master Kit (Roche Diagnostics, France) for viruses. A C1000 Touch thermal cycler (Bio-rad, Hercules, CA, USA) was used to perform each. Negative controls (mix mix) and positive controls (DNA from a bacterial or viral strain) were included in all runs. Positive pathogen amplification results were evaluated against a cycle threshold (CT) value of ≤ 35.

Statistical analysis

Stata 17.0 software was used for statistical analysis. Categorical variables were presented as percentages, while continuous variables were presented as median and interquartile. The study was not formally powered to detect differences in pathogen carriage or co-carriage per year; analyses were primarily descriptive, aiming to explore overall trends across the study period. Our main outcomes were the presence of respiratory symptoms among pilgrims and the fact of consulting for such symptoms. To improve the robustness of the model and reduce instability due to sparse data, variables with a prevalence of less than 5% in the study population were excluded from the risk factor analysis. Co-carriage was defined as the detection of two or more pathogens in the same sample. It was included as an independent variable in descriptive summaries and multivariable analyses to assess its association with clinical symptoms and consultation behavior.

Logistic regression models were employed to identify factors associated with two primary outcomes: the presence of respiratory symptoms (symptomatic vs. asymptomatic) and the severity of disease (severe vs. non-severe). Severity of disease was defined as consulting for respiratory symptoms. Univariable logistic regression was first conducted to estimate crude odds ratios (ORs) and corresponding 95% confidence intervals (CIs).

Two multivariable models were constructed for each outcome. Model 1 included all positive respiratory pathogens without adjustment, while model 2 was adjusted for age and gender to account for potential confounding. Statistical significance was defined as a two-sided p-value < 0.05.

Results

Characteristics of the study population

The study included 1067 pilgrims (2017–2023) distributed as follows: 328 healthy controls, 298 pilgrims with respiratory symptom who did not consult and 441 consulting patients who consulted for respiratory symptoms (438 Mbacké patients and 3 cohort pilgrims). 6.1% cohort pilgrims declined to participate. 6.5% of cohort pilgrims were included in at least two years of study. The attack rate of respiratory symptoms in cohort pilgrims on return was 48%. Healthy controls were younger (27 years) than non-consulting ill pilgrims (23 years) and consulting patients (18 years), and more likely to be male (54% vs. 42% and 41%). Vaccination rates against respiratory infections were very low, overall (Table 1).

Table 1.

Demographics of the study population and vaccination status.

Variables		Healthy controls	Non-consulting ill pilgrims	Consulting patients
		N = 328 (1)	N = 298 (2)	N = 441 (3)
Year	2017	64 (19.5%)	34 (11.4%)	−
	2018	52 (15.9%)	37 (12.4%)	52 (11.8%)
	2019	23 (7.0%)	56 (18.8%)	112 (25.4%)
	2020	46 (14.0%)	42 (14.1%)	109 (24.7%)
	2021	57 (17.4%)	51 (17.1%)	53 (12.0%)
	2022	44 (13.4%)	45 (15.1%)	57 (12.9%)
	2023	42 (12.8%)	33 (11.1%)	58 (13.2%)
Mean age (years)		27.1 ± 17.0	22.8 ± 15.0	18.2 ± 18.9
Age class (years)	0–15	83 (25.4%)	99 (33.3%)	232 (53.8%)
	16–45	199 (60.9%)	161 (54.3%)	150 (34.8%)
	> 45	45 (13.7%)	37 (12.4%)	49 (11.4%)
Missing data		1	1	10
Females		151 (46.0%)	174 (58.4%)	258 (58.8%)
Males		177 (54.0%)	124 (41.6%)	181 (41.2%)
Missing data		−	−	2
Vaccination against COVID-19 (2021–2023)		9 (2.7%)	14 (4.7%)	17 (3.9%)
Vaccination against invasive pneumococcal		13 (4.0%)	18 (6.0%)	−
Vaccination against influenza		1 (0.3%)	0 (0.0%)	−

Respiratory symptoms and pathogen carriage

Consulting patients were more likely to present with cough and fever, while non-consulting ill pilgrims were more likely to report dyspnea and sore throat. Rhinitis was frequent in both groups. 28 (6.3%) and 3 (0.7%) consulting patients and non-consulting ill pilgrims, respectively, were hospitalized. The majority of consulting patients received antibiotics (71%) while non-consulting ill patients rarely did so (3%) (Table 2).

Table 2.

Prevalence of clinical symptoms.

Variables	Non-consulting ill pilgrims (N = 298)	Consulting patients (N = 441)	P-value
Cough	202 (67.8%)	429 (97.3%)	< 0.0001
Dyspnea	41 (13.8%)	15 (3.4%)	< 0.0001
Sore throat	85 (28.5%)	85 (19.3%)	0.003
Rhinitis	199 (66.8%)	302 (68.5%)	0.63
Fever	96 (32.2%)	186 (42.2%)	0.006
Antibiotic use	8 (2.7%)	311 (70.5%)	< 0.0001

Results on pathogen carriage are presented in Tables 3 and 4. Among healthy controls, 14% of respiratory samples tested positive for at least one virus, with rhinovirus (7%) and endemic coronaviruses (4%) the most frequent, and 74% tested positive for bacteria, with Haemophilus spp. (45%) and S. aureus (29%) the most frequent. S. pneumoniae and K. pneumoniae carriage rates were about 20% and 19% respectively. Among non-consulting ill pilgrims, viral carriage was higher (28%), with again rhinovirus (14%) and endemic coronaviruses (8%) accounting for most cases. Influenza and respiratory syncytial virus carriage were rare. Bacterial carriage in non-consulting ill pilgrims was in the same range of that of healthy controls (79%), with again Haemophilus spp. (45%) and S. aureus (31%) the most frequent. However, S. pneumoniae (32%) carriage rate was higher than in healthy controls, while that of K. pneumoniae remained in the same range (20%). Among consulting patients, 62% tested positive for at least one virus, representing twice the prevalence of viral carriage in non-consulting ill pilgrims, with influenza A (29%), respiratory virus syncytial (14%) and rhinovirus (9%) the most frequent. Endemic coronaviruses were rarely detected (2%). A high rate of bacterial carriage was observed (82%), notably due to Haemophilus spp. (62%), S. pneumoniae (47%) and M. catarrhalis (39%), for which prevalence was higher than in the non-consulting ill pilgrim group. By contrast, K. pneumoniae carriage was very low in consulting patients (1%). Co-infection with virus and bacteria was 12% in healthy controls, 23% in non-consulting ill pilgrims and 51% in consulting patients.

Table 3.

Associated factors with presence of respiratory symptom.

Associated factors	No symptom	Symptom	Univariable analysis		Multivariable analysis (Model 1)		Multivariable analysis (Model 2)*
	n (%)	n (%)	Crude OR [95%CI]	p-value	OR [95%CI]	p-value	Adjusted OR [95%CI]	p-value
Age
0–15	83 (20.0)	331 (80.0)	Reference				Reference
16–45	199 (39.0)	311 (61.0)	0.39 [0.29–0.53]	< 0.001			0.73 [0.51–1.03]	0.08
> 45	45 (34.4)	86 (65.6)	0.48 [0.31–0.74]	0.001			0.90 [0.55–1.48]	0.69
Male gender	176 (36.6)	305 (63.4)	0.61 [0.47–0.79]	< 0.001			0.58 [0.43–0.77]	< 0.001
Influenza A	1 (0.7)	139 (99.3)	75.75 [10.55–544.08]	< 0.001	53.97 [7.41–393.18]	< 0.001	53.34 [7.30–389.42]	< 0.001
Influenza B	1 (3.8)	25 (96.2)
Respiratory syncytial virus	0 (0)	66 (100)
Metapneumovirus	0 (0)	13 (100)
Adenovirus	4 (25.0)	12 (75.0)
Endemic coronaviruses	14 (29.8)	33 (70.2)
SARS-CoV-2 (N = 741)	5 (35.7)	9 (64.3)
Virus co-infection	2 (7.1)	26 (92.9)
Human rhinovirus	22 (21.4)	81 (78.6)	1.71 [1.05–2.80]	0.03	1.01 [0.54–1.89]	0.98	1.13 [0.60–2.12]	0.70
Haemophilus spp.	149 (26.3)	418 (73.7)	1.56 [1.20–2.03]	0.001	1.03 [0.72–1.47]	0.87	1.03 [0.72–1.49]	0.86
S. pneumoniae	64 (17.5)	301 (82.5)	2.83 [2.08–3.86]	< 0.001	1.90 [1.22–2.94]	0.004	1.90 [1.22–2.98]	0.01
S. aureus	94 (32.6)	194 (67.4)	0.89 [0.66–1.18]	0.41	0.94 [0.63–1.40]	0.75	0.94 [0.63–1.42]	0.78
K. pneumoniae	62 (48.8)	65 (51.2)	0.41 [0.28–0.60]	< 0.001	0.53 [0.34–0.83]	0.006	0.58 [0.36–0.92]	0.02
M. catarrhalis	45 (16.4)	230 (83.6)	2.84 [2.01–4.04]	< 0.001	1.77 [1.14–2.75]	0.01	1.62 [1.03–2.55]	0.04
B. pertussis	0 (0)	1 (100)
M. pneumoniae	0 (0)	0 (0)
Bacteria co-infection	121 (24.0)	383 (76.0)	1.84 [1.41–2.40]	< 0.001	1.14 [0.65–2.01]	0.64	1.10 [0.62–1.96]	0.74
Virus-bacteria co-infection	40 (12.0)	294 (88.0)	4.76 [3.31–6.83]	< 0.001	2.69 [1.69–4.29]	< 0.001	2.61 [1.62–4.20]	< 0.001

*Model 2: adjusted for age and gender of pilgrims.

Influenza B, respiratory syncytial virus, metapneumovirus, adenovirus, human coronaviruses, SARS-CoV-2, virus co-infection, Bordetella pertussis, and Mycoplasma pneumoniae were not included in the analysis because of low prevalence.

Table 4.

Associated factors with severity of disease (consulting for respiratory symptom).

Associated factors	No-severe disease	Severe disease	Univariable analysis		Multivariable analysis (Model 1)			Multivariable analysis (Model 2)*
	n (%)	n (%)	Crude OR [95%CI]	p-value	OR [95%CI]		p-value	Adjusted OR [95%CI]		p-value
Age
0–15	99 (29.9)	232 (70.1)	Reference					Reference
16–45	161 (51.8)	150 (48.2)	0.40 [0.29–0.55]	< 0.001				0.64 [0.42–0.97]		0.04
> 45	37 (43.0)	49 (57.0)	0.57 [0.35–0.92]	0.02				1.34 [0.72–2.49]		0.36
Male gender	124 (40.7)	181 (59.3)	0.98 [0.73–1.33]	0.92				0.89 [0.62–1.28]		0.54
Influenza A	10 (7.2)	129 (92.8)	11.91 [6.14–23.11]	< 0.001	7.81 [3.80–16.04]		< 0.001	8.08 [3.90–16.73]		< 0.001
Influenza B	1 (4.0)	24 (96.0)
Respiratory syncytial virus	3 (4.6)	63 (95.4)
Metapneumovirus	2 (15.4)	11 (84.6)
Adenovirus	4 (33.3)	8 (66.7)
Endemic coronaviruses	24 (72.7)	9 (27.3)
SARS-CoV-2 (N = 413)	6 (66.7)	3 (33.3)
Virus co-infection	10 (38.5)	16 (61.5)
Human rhinovirus	42 (51.8)	39 (48.2)	0.59 [0.37–0.94]	0.03	0.40 [0.21–0.75]		0.004	0.40 [0.21–0.75]		0.004
Haemophilus spp.	146 (34.9)	272 (65.1)	1.68 [1.24–2.26]	0.001	1.40 [0.88–2.20]		0.15	1.48 [0.93–2.38]		0.10
S. pneumoniae	96 (31.9)	205 (68.1)	1.83 [1.35–2.48]	< 0.001	1.50 [0.89–2.52]		0.13	1.50 [0.88–2.55]		0.14
S. aureus	93 (47.9)	101 (52.1)	0.65 [0.47–0.91]	0.01	0.93 [0.58–1.50]		0.78	0.92 [0.56–1.48]		0.72
K. pneumoniae	60 (92.3)	5 (7.7)	0.04 [0.02–0.11]	< 0.001	0.07 [0.03–0.19]		< 0.001	0.07 [0.03–0.19]		< 0.001
M. catarrhalis	57 (24.8)	173 (75.2)	2.73 [1.93–3.86]	< 0.001	2.82 [1.75–4.53]		< 0.001	2.57 [1.56–4.22]		< 0.001
B. pertussis	0 (0)	1 (100)
M. pneumoniae	0 (0)	0 (0)
Bacteria co-infection	147 (38.4)	236 (61.6)	1.18 [0.88–1.59]	0.26	0.58 [0.29–1.17]		0.13	0.54 [0.26–1.11]		0.10
Virus-bacteria co-infection	69 (23.5)	225 (76.5)	3.46 [2.49–4.80]	< 0.001	2.85 [1.77–4.59]		< 0.001	2.83 [1.74–4.61]		< 0.001

*Model 2: adjusted for age and gender of pilgrims.

Influenza B, respiratory syncytial virus, metapneumovirus, adenovirus, endemic coronaviruses, SARS-CoV-2, virus co-infection, Bordetella pertussis, and Mycoplasma pneumoniae were not included in the analysis because of low prevalence.

Risk factors for respiratory symptoms and consulting for such symptoms

In the multivariable model adjusted for age and gender (model 2), male gender was significantly associated with a lower risk of respiratory symptoms (aOR 0.58; 95% CI 0.43–0.77; p < 0.001). Age was not significantly associated with symptom presence after adjustment.

Several pathogens were independently associated with the presence of symptoms. Influenza A infection showed a strong positive association (aOR 53.34; 95% CI 7.30–389.42; p < 0.001). Infection with S. pneumoniae (aOR 1.90; 95% CI 1.22–2.98; p = 0.01), M. catarrhalis (aOR 1.62; 95% CI 1.03–2.55; p = 0.04), and virus-bacteria co-infection (aOR 2.61; 95% CI 1.62–4.20; p < 0.001) were all significantly associated with increased odds of symptom presence. In contrast, infection with K. pneumoniae was associated with decreased odds of reporting symptoms (aOR 0.58; 95% CI 0.36–0.92; p = 0.02). No significant associations were observed for Haemophilus influenzae, S. aureus, or human rhinovirus after adjustment.

Regarding the severity of disease, defined as consulting for respiratory symptom, participants aged 16–45 years had significantly lower odds of severe disease compared to those aged 0–15 years (aOR 0.64; 95% CI 0.42–0.97; p = 0.04), while gender was not significantly associated with severity.

Influenza A infection remained a strong associated factor of severe disease (aOR 8.08; 95% CI 3.90–16.73; p < 0.001). M. catarrhalis infection (aOR 2.57; 95% CI 1.56–4.22; p < 0.001) and virus-bacteria co-infection (aOR 2.83; 95% CI 1.74–4.61; p < 0.001) were also independently associated with increased severity. Conversely, infection with K. pneumoniae (aOR 0.07; 95% CI 0.03–0.19; p < 0.001) and human rhinovirus (aOR 0.40; 95% CI 0.21–0.75; p = 0.004) were negatively associated with severe disease. No significant associations were observed for Haemophilus spp., S. pneumoniae, S. aureus, or bacterial co-infection.

Discussion

Mass gatherings can facilitate the transmission of microbial pathogens, resulting in both acquisition of pathogens and increased prevalence of symptomatic infections. This has been well exemplified during the Hajj Muslim pilgrimage in Saudi Arabia. Up to 90% of pilgrims may present with respiratory symptoms and high rates of acquisition of several respiratory viruses and bacteria have been observed in link with crowding conditions⁴. It is, however, difficult to formally establish a relationship between respiratory carriage of potential pathogens and clinical symptoms, since asymptomatic carriage may occur, and mixed infections are frequent. Asyptomatic carriage of respiratory viruses is now a well recognized phenomenon at least for influenza virus, SARS-CoV-2 and rhinovirus^5–7. Like Hajj, the GMT exposes participants to overcrowding, although it takes place in a different geographical context and involves a much younger population with limited access to healthcare facilities during the event⁵. High rates of viruses and bacteria were noted in ill pilgrims returning from Touba, including rhinovirus, endemic coronaviruses, Haemophilus spp., S. pneumoniae and S. aureus. These results are consistent with those observed in French Hajj pilgrims and Hajj pilgrims of other nationalities, using the same detection methods^3,9–15.

In a previous study conducted from 2017 to 2021 on cohorts of GMT pilgrims, most of whom did not consult for their symptoms, we concluded that female pilgrims were more at risk for respiratory symptoms and that overall virus acquisition and acquisition of S. pneumoniae were independently associated with respiratory symptoms². In this work, we extended the study period to 2022–2023 and also used data obtained in patients consulting for respiratory symptoms during the GMT. This increased the number of subjects from 535 to 1067, giving more statistical power to our analysis. Also, instead of using pathogen acquisition as a variable, as in our previous work², we used pathogen carriage at the time of returning home in cohort pilgrims and at the time of symptoms in Mbacké patients, in order to be more consistent between these two groups of participants. In addition, we addressed the effect of viral-bacterial co-infection on symptoms. Here we confirm the higher likelihood of female GMT participants to present with respiratory symptoms and our results suggest a possible role of S. pneumoniae as a pathogen responsible for respiratory symptoms in the context of the GMT. In addition, our results suggest a possible role of co-infections with viruses and bacteria, and the specific role of M. catharralis and particularly that of influenza A virus in the occurrence of respiratory symptoms. These new findings likely result from the increased number of participants included in our current analysis and also from the inclusion of Mbacké patients. Indeed, the proportion of ill cohort pilgrims with influenza carriage was very low in comparison to that of Mbacké patients. We believe that this difference was due to the timing of sampling that was done 6 to 9 days after return in cohort pilgrims, when participants may have cleared their influenza virus infection, while Mbacké patients were all sampled at the time of symptoms. It is very likely that influenza B virus, respiratory syncytial virus and M. catharralis carriage was also partially cleared in returned non-consulting cohort pilgrims, since prevalence in consulting patients was clearly higher. Nevertheless, we cannot completely exclude that ill pilgrims who consulted were more severely ill that those who did not, since cough and fever in the latter group was more frequent, which could possibly indicate a higher prevalence of lower respiratory tract infections. Conversely, sore throat, which is indicative of upper respiratory tract infection, was more frequent in the non-consulting group. The presence of dyspnea in higher number of non-consulting patients is difficult to understand. Difficulty in breathing is associated with severe disease which would require medical consultation or hospitalization. The markedly lower rates of K. pneumoniae and endemic coronavirus carriage in consulting patients might reflect lower circulation of these microbes in the Mbacké urban environment, as compared to cohort pilgrims living in a rural setting.

Our study has some limitations. Inclusions were based on pilgrims from two villages and patients consulting at a single healthcare center. Therefore, the sample may not be representative of all GMT participants. Cohort pilgrims were enrolled 6–9 days after their return. In some individuals, symptoms will subside during this period and although they may have respiratory symptoms 5–6 days back they may be asymptomatic at the time of sample collection. The consulting group is drawn from a medical post, whereas the other two are recruited by the study team. This introduces potential selection bias, particularly if those seeking care differ systematically (in severity, age, comorbidities) from non-consulting participants. Also, in consulting patients, there was no assessment of prior influenza or pneumococcal vaccination, which could confound carriage dynamics.

We only included consulting patients seen during the daytime, and we have no information about their follow-up after leaving the healthcare center. qPCR cannot distinguish between active and inactive pathogens. Also, qPCR detects DNA/RNA, which does not distinguish colonization from active infection, especially for bacteria such as S. pneumoniae, Haemophilus spp., and S. aureus, which are common colonizers. Therefore, without data on pathogen load, clinical severity or inflammatory markers, it is not possible to formally infer that carriage equates to causality.

Despite its limitations, our work provides a comprehensive description of respiratory symptoms experienced by pilgrims participating in the GMT, as well as microbiological data, suggesting the pathogenic role of some of these microorganisms, and identifies certain risk factors for respiratory symptoms. Based on these results, GMT pilgrims should be advised to wear masks in enclosed spaces, as this has been shown to be effective against respiratory symptoms in the context of the event² and demonstrated in other studies of large gatherings such as Hajj^16,17. Given the very low rates of vaccination against influenza and invasive pneumococcal infections among GMT pilgrims, pilgrims with comorbidities should be encouraged to get vaccinated against these diseases, as recommended during Hajj¹⁸. Vaccination and monoclonal antibodies are now market-approved to protect infants¹⁹ and these tools may be considered in the future for GMT participants. We did not find a major role for SARS-CoV-2 as responsible for respiratory tract infections at the GMT on which to base recommendations for systematic vaccination of pilgrims against COVID-19. Hand-washing with soap and water should be encouraged, as it has proved to be an effective means of combating respiratory viruses at the GMT². Finally, because female pilgrims are more at risk for respiratory tract infections, they should be specially targeted for information campaigns around preventive measures. Because they are usually in charge of childcare, this would also benefit younger participants. Regarding the treatment of respiratory tract infections at the GMT, in the absence of point-of-care PCR diagnostic tools, rapid influenza diagnostic tests should be systematically used and oseltamivir proposed to positive patients when available. If judged necessary, first-line antibiotherapy should target molecules active on both S. pneumoniae and M. catarrhalis, like amoxicillin-clavulanic acid or second and third-generation cephalosporin.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

C.D. writing original draft, collected samples, wrote the manuscript, conducted the qPCR technique. I.O. collected samples. V.T.H. did the statistical analysis, reviewed the manuscript. M.S., H.B. contributed to the experimental design. N.G, D.N. collected samples, conducted the qPCR technique. P.G. Writing original draft, contributed to the experimental design, wrote the manuscript, coordinated the work. C.S. contributed to the experimental design and coordinated the work.

Funding

This study was supported by the Institut Hospitalo-Universitaire (IHU) Méditerranée Infection, and the French National Research Agency under the ‘Investissements d’Avenir’ program, reference ANR-10-IAHU-03.

Data availability

Data is provided within the manuscript.

Declarations

Consent for publication

All authors contributed to and approved the current version of the manuscript.

Competing interests

The authors declare no competing interests.

Ethics

The protocol was approved by the National Ethics Committee forHealth Research in Senegal (SEN17/62) and performed in accordance withthe good clinical practices recommended by the Declaration of Helsinki andits amendments.