Outbreak of diarrhoeal illness in participants in an obstacle adventure race, Alpes-Maritimes, France, June 2015

C Six 1 , S Aboukais 2 , S Giron 1 , J D’Oliveira 3 , F Peloux-Petiot 2 , F Franke 1 , H Terrien 2 , F Dassonville 2 , J Deniau 1 , K AmbertBalay 4 , T Chesnot 5 , R Ruimy 6 , M Pélandakis 7 , P Basset 8 , M Munoz Rivero 2 , P Malfait 1 1. Santé publique France, French national public health agency, Regional unit (Cire Provence-Alpes-Côte d’Azur and Corsica), Marseilles, France 2. Regional Health Agency of Provence-Alpes-Côte d’Azur, Marseilles, France 3. Laboratory ‘Santé Environnement Hygiène’ (CARSO), Lyon, France 4. National Reference Center for Enteric virus, Dijon, France 5. Laboratory for Hydrology of the French Agency for Food, Environmental and Occupational Health & Safety, Nancy, France 6. University Hospital Laboratory of Bacteriology, Nice, France 7. University of Lyon 1, Faculty of Pharmacy, Villeurbanne, France 8. Dokever, Paris, France


drology
Frenc
Agency for Food, Environmental and Occupational Health & Safety
NancyFrance


University Hospital Laboratory of Bacteriology
NiceFrance


Faculty of Pharmacy
University of Lyon 1
VilleurbanneFrance


Dokever, ParisFrance

Outbreak of diarrhoeal illness in participant in an obstacle adventure race, Alpes-Maritimes, France, June 2015
3777901396B13310A950F843B8E57A3D10.2807/1560-7917.ES.2016.21.23.30253Article submitted on 08 September 2015 / accepted on 17 February 2016 / published on 09 June 2016
An acute gastroenteritis (AG) outbreak occurred among participants in an obstacle race in France in the summer of 2015.An investigation in two phases was conducted to identify the source of infection and document the extent of the outbreak.First, a message on a social media website asked racers to report any symptoms by email to the Regional Health Agency of Provence-Alpes-Côte d'Azur.Second, a retrospective cross-sectional study was conducted through an interactive questionnaire for all participants, followed by an analytical study of potential risks factors.Of 8,229 persons registered, 1,264 adults reported AG resolved within 48 hours.Of adults who reported AG, 866 met the case definition.Age group, departure time and ingestion of mud were associated with AG.Twenty stool specimens tested negative for bacteria.All four stool samples tested for viruses were positive for norovirus genogroup I and genotype 2. No indicator bacteria for faecal contamination were found in drinking water but muddy water of ponds tested positive.The outbreak was possibly caused by human-to-human transmission of a norovirus introduced by one or more persons and transmitted through contaminated mud.Risks related to similar races should be assessed and recommendations be proposed to raise awareness among health authorities and organisers.

Introduction

Obstacle races are extreme sport events, combining a difficult path and obstacles.On a distance of several kilometres, obstacles such as walls, nets, ice baths, barbed w res, mud mounds, tunnel crawls, dark rooms, live electrical wires are positioned along the path.These sport events have been growing in popularity during the last years.They gather several thousands of participants and are now a worldwide phenomenon.

Racers are heavily exposed to mud and water.They frequently fall, face first, into mud or have their heads submerged in water.Sometimes, there is a different path available for children.

ents are
usually held in rural areas and often include man-made slurry fields (a mixture of soil and water).In areas commonly frequented by animals, topsoil used to make slurry fields can be con

, in the south of
France.On a social media website, many other participants reported also suffering from similar symptoms.An investigation was immediately conducted in order to identify the source of infection and document the extent of the outbr ak.


Methods

The ARS PACA, responsible for the outbreak investigation and implementation of control measures, requested support of the Regional Office of the French Public Health Agency (Cire Sud).


Epidem

logical investigation

The inv
stigation of the outbreak was conducted in two phases.The first phase started immediately, upon receipt of the outbreak alert, on 22 June 2015.ARS PACA and Cire Sud sent a message on a social media website informing the participants of the obstacle race that an investigation was conducted and asked them to

eport any recent or current gastrointe
tinal illness by email to a dedicated address of ARS.This information was also relayed by the organisers of the race and the local press.

The second phase, aimed at collecting additional information, was conducted retrospectively through an interactive application, Voozano (a secure web-based platform for hosting personal health data) including sections on socio-demographic characteristics, presence of symptoms or not, and sections on potential exposures (catering, means of contact with mud).This short questionnaire (available from the authors upon request) was intended for all participants of the obstacle race, whether or not they had clinical signs of AG.On 30 June, the company that had organised the event sent the Internet link for this survey by email in a newsletter to all the participants of the obstacle race.

Cross-sectional descriptive studies were conducted in two phases, and were followed by an analytical study of the potential risks fac

second phase).


Case de
initions

A case of AG was defined as a racer in the obstacle race that took place on 20 June, with self-reported gastrointestinal illness (vomiting and/or diarrhoea) associated or not with other symptoms within eight days from the race.

A secondary case of AG was defined as a person who did not participate in the obstacle race as racer, with gastrointestinal illness following at least 24 hours after a contact with a case of AG among racers.


Microbiological investigation

We recommended all laborator es to test stool specimens from the racers for Salmonella, Shigella, Campylobacter and Yersinia, and to send stool specimens to the National Reference Laboratory (NRL) for Enteric viruses in Dijon in order to test for norovirus, sapovirus, rotavirus, adenovirus and astrovirus Food and environmental investigations

The company organising the event was contacted on 23 June to collect information about the race (organisation, description of the obstacles, digging works, etc.)   Self-reported data.

and the catering company

to obtain information about me
us and food samples for analysis.

Samples of muddy water were collected on 23 June from the ponds of the race and samples of drinking water from the city distribution system and tested for bacteria indicating faecal contamination (Escherichia coli, total coliforms, Enterococci, aerobic microorganisms) and for Salmonella, enterovirus, Vibrio cholerae and Vibrio parahaemolyticus, and free-living amoebae We provided recommendations on measures to prevent secondary spread d ring a press conference organised on 22 June, through a weekly report of the Cire Sud, on the ARS PACA website, and we posted prevention messages on a social media website.The outcome of the investiga ion was also published on all these websites.


Results


Description of the event

The obstacle race took place over one day.A total of 8,229 participants were registered: 7,804 adults and 425 children (aged 7-12 years).Adult racers completed 13 km and 22 obstacles.Departures took place every 20 minutes, from 9:00 to 16:00, by waves of around 400 participants.On the path, dunes had been constructed and artificial ponds dug and filled with water from

e city distribution sys
em mixed with the original soil.The children's path was shorter and did not include ponds of muddy water.

Three feed stations were set up along the race course.Food included pre-wrapped cakes, pre-cut bananas, dried apricots, all served in plastic containers from which participants picked them with their bare hands.Drinks included bottled water, energy drinks and beer in cups.

The race was held near a riding centre, and some participants reported spontaneously seeing horse dung on or near the track, or smelling sewage.


Epidemiological investigation

Description of the outbreak

hase 1 of the epidemiological i
vestigation was conducted from 22 to 27 June; we received 1,370 emails, 81% (1,111/1,370) of which had been sent within 24 hours after the message about the outbreak investigation was published on the social media website.Among the respondents, 1,300 AG cases were reported: 1,264 ad

ts (attack rate: 16
; 1,264/7,804) and 36 secondary cases (not included in this analysis).

Phase 2 was conducted from 30 June to 27 July and a total of 748 questionnaires were completed (for 745 adults and 3 children).Among the participants who filled out the questionnaire, more than half (404/748; 54%) did it the day the newsletter came out and one third (247/748; 33%) the day after.There were 375 AG cases and

3 non cases.Secondary cases
eported by respondents rose to a total of 177 cases, not included in this analysis.

The mean age of adult respondents was 33 years (range: 18-61 years).The male/female sex ratio was 1.2.Participants came mostly from Alpes-Maritimes (580/745; 78%), followed by other French departments or other countries.

In both phases, no cases were reported among the child racers.Only adults were included in the analysis.

The median incubation period was one day (ranges: 0-4 days for phase 1 and 0-5 days for phase 2).

Self-reported onset of AG symptoms was available for 820 participants (phase 1) and 372 (phase 2). ) Of the symptoms 97% (792/820) and 95% (354/372) occurred on day 1 and 2 after the race for phase 1 and 2, respectively (Figure 1).

In phase 1, the description of symptoms was available for 1,001 (79%) of The departure time was available for 287 AG cases in phase 1 and

0 in phase
(Figure 2).Cases were identified for all departure waves, except one at 15:40 for phase 1.The number of cases was low for the first waves (range: 1-17 cases), and the highest for the fifth wave, with 48 (phase 1) and 49 AG cases (phase 2).The curves of AG cases over time for the two phases of the investigation follow the same dynamic (trend chisquare not significant).


Study of risks factors

Age group, departure time and ingestion of mud were associated with AG at the p < 0.2 level in the univariate analysis (Table 1).

The final multivariable model included age group, departure time and ingestion of mud (Table 2).Younger participants (aged 18-27) years had a significantly higher risk or AG in the multivariate analysis compared with older participants.The risk of AG was higher in the morning, from 10:00 to 14:00 and still high in the afternoon compared with the first hour of departur waves (9:00-10:00).The ingestion of mud is also a significant risk factor of AG.


Microbiological investigations

We obtained the results of 20 stool samples.Results of bacteriological testing were negative for all samples, except one (positive for Shigella sonnei).Among the four specimens sent to the NRL for enteric viruses, all were positive for norovirus genogroup I and genotype 2 (GI.2).


Food investigation

During the race, the participants did not carry their own food and drinks.However, they consumed beverages and food distributed in feed stations.One third of the cases reported having eaten food sold by the catering company after the ra e.No AG case was reported among non-participants, organisers and staff of the event.The catering company had no food samples for microbiological analysis.


Environmental investigation

Drinking-water from the city distribution system and environmental water samples from muddy water ponds were collected on 23 June.All three drinking water samples were negative for the indicator bacteria (total coliforms, E. coli, Enterococcus), viruses and parasites.Of the environmental water specimens (n=5) taken from five ponds on 23 June, all were found contaminated by aerobic microorganisms at 36 °C (with counts ranging from 86 to 3,200,000 UFC/100 mL) and 22 °C (with counts ranging from 2,500 to 1,400,000 UFC/100 mL), indicating bacterial contamination.All specimens were negative for Salmonella, Campylobacter, Enterovirus, V. cholerae and V. parahemolyticus.All environmental samples were positive for Naegleria spp.(with cells ranging from 400 to 280,000/L), but the pathogenic species N. fowleri was not detected.Norovirus genogroup I and II were not detected.


Discussion

A large AG outbreak occurred among the participants of obstacle race in the department of Alpes-Maritimes, France, on 20 June 2015.Of 7,804 adult participants, 1,264 were ill and 866 met the case definition of AG.This outbreak occurred from 20 to 25 June and the epidemiological curves during two different phases of the outbreak investigation were characteristic of a point-source outbreak.Moreover, they were similar and showed a higher number of cases on day 1 and 2 after the race.The clinical characteristics reported, the incubation period of the AG, the presence of vomiting, the resol tion of symptoms for most cases within 48 hours and the absence of severe clinical forms, were compatible with a viral origin of the outbreak.

The epidemiological investigation enabled us to identify mud ingestion as the main risk factor of developing AG; no other source of infection was identified.

Several arguments support a common source outbreak of human origin related to the ingestion of contaminated mud.In outbreaks such as the one described here, the first hypothesis usually considers food sources.However, the first information r ported by the participants did not support a food source, because of the diversity of supplies (feed stations only for participants in the race, food stalls for everyone, food brought by racers).Information gathered from organisers showed that no case of AG had been reported among the 360 members of the organisation and volunteers who received the food which was also sold at food stalls open to the public.

Unfortunately, no food samples were available for labora ory testing.Among participants, 90% of cases and 89% of non-cases (phase 2) consumed food from mass catering served at food stalls.However, the hypothesis that participants could have contaminated one another when dipping their muddy hands into the plastic containers with pre-cut bananas or dried fruits seems plausible.Another argument for contaminated mud as vehicle in the outbreak is that no child has been reported ill.The children's path was different from the adults' and did not have any obstacle with muddy water.This reinforces the idea of a human contamination in the obstacles with muddy water of the adult path.Secondary spread between ill participants and their relatives happened outside of the context of the race.Norovirus GI.2 was identified in the four stool samples.This strain of human origin [11], was possibly introduced by one person or more of the team preparing the path or by participants in the first waves.In spite of the negative tests for norovirus in the muddy water, we cannot exclude that it was contaminated with such viruses [12].Indeed, the sensitivity of the norovirus detection method could have been significantly reduced because only low volumes of muddy water were available for testing.Norovirus can persist in the environment, but the viral contam nation of the water may have been partly reduced over the three days that preceded the sampling.Last, it is possible that the most contaminated ponds may have not been sampled.We consider it highly probable that the virus ad been disseminated in the days preceding the race or early during the race [13].

Although our investigations led us to identify a point source with probable norovirus contamination as cause of the outbreak, there are some difficulties and limitations.The phase 1 of the i vestigation had to be conducted quickly, but it would have been better to have the time to ask the participants a few specific questions instead of just asking them to report any AG.This would have led to improved data quality.Nevertheless, participants gave us spontaneously detailed descriptions of their symptoms which allowed us to use the data.

Another limitation is that it was not possible to identify the exact number of cases.Phase 1 listed the participants who reported their illness voluntarily.More than 1,000 replies were sent by email within the 4 hours that followed the request for information from participants in the race.But there is a possibil ty that a greater number of persons was aware of the message but did not take the time to report their illness, in particular if symptoms were mild.

During phase 2, the response rate was lower, which prevented us from estimating the actual number of cases.As the event company had email addresses of participants, it had been agreed with the organisers to go through them to ask the participants to fill out the questionnaire, at the Internet link provided.The information was sent to the event company on 25 June, but was only inserted in a newsletter sent to the participants on 30 June i.e. 10 days after the event and eight days after he peak of the outbreak.It is likely that this newsletter was seen by more participants than those who replied during phase 1, because the message was sent directly to their email address.However, this email did not trigger as many as replies as expected, maybe because the illness was not th