A Review of the Evidence provided by the Welsh Technical Advisory Group for Prolonged Shutting of Gyms in Wales.
The Welsh Government Technical Advisory Group (TAG) published a paper titled –‘Environmental Science Subgroup SARS-CoV-2 infection risks at indoor exercise facilities’ on 8th January 2021 identifying the risks associated with the use of indoor exercise facilities defined as gyms, leisure centres and dance or exercise classes within the context of COVID-19. The TAG report identifies a high potential for superspreading events (SSE) within indoor exercise facilities as the ‘chief’ reason for the long-term closure and delayed reopening of facilities. In section 1 of the TAG report they use an investigation into South Korean Zumba classes (Jang et al., 2020) and an ‘informal database’ of worldwide SSE (Swinkels, 2020) as the two main sources of evidence for the high risk of SSE in gyms. However, both sources of evidence require further scrutiny, as the quality and robustness of the research and its interpretation by TAG is questionable and the following report briefly highlights these issues.
Mistakes in data entry in the informal database of worldwide SSE
The TAG report highlights an ‘informal database’ developed by Swinkels, (2020) of SSE as evidence for shutting gyms. A summary table of the specific SSE in gyms, dance and fitness classes identified in the informal database can be seen at the bottom of this document. The TAG report states,
“A rapid review of reported superspreading events in an informal database ( https://kmswinkels.medium.com/covid-19-superspreading-eventsdatabase-4c0a7aa2342b ) comprising 2,007 events to the 27th of November 2020 identified 615 cases from 15 events (on average, 41 cases per event) associated with dance/fitness classes (9 events) or gym use (6 events).”
Despite the TAG report claiming 6 SSE were associated with gyms on the database, 3 of the 6 reported SSE occurred at dance or spin studios with errors at data entry, namely they were recorded as gyms but should have been recorded as dance or spin classes and 1 case of the 6 recorded as gym is a duplicate entry. The 6 SSE supposedly occurring in gyms from the database are summarised below. (Database id codes are in brackets)
1. An outbreak of 65 cases in Cheonan, South Korean (KOR01) and investigated by the Korean Disease Control and Prevention Agency is recorded as gym, but is actually at a Zumba dance class. This is discussed in depth below.
2. An outbreak of 80 cases in Hamilton, Canada (CAN26) is recorded as a gym, but was a standalone spin studio whereby group cycle classes were performed in a small enclosed room with up to 40 participants.
3. An outbreak of 18 cases at Elmhurst, Illinois, United States is recorded as a gym, but occurred at an indoor group fitness class. This is a group fitness studio and the company website states they provide a ‘total-body group workout…in our studio’.
4. An outbreak of 5 cases at Aichi, Japan in February 2020 was in a gym, although there was no indication of precautionary measures taken and the number of cases is highlighted as ‘uncertain’ on the database. Is 5 cases a substantial SSE?
5. An outbreak of 40 cases at a fitness facility in Rayleigh, North Carolina in November 2020, provides little information about the details of the outbreak. The facility provides up to 17 group fitness classes per day and the newspaper report that the facility ‘had 50-60 people in classes before the pandemic began. Now, they have 35 people or fewer per class’. The database suggests that no social distancing or disinfecting mitigation measures were followed.
6. A duplicate copy in the database of the outbreak of the 40 cases at a fitness facility in Rayleigh, North Carolina (point 5). It is recorded to have taken place on 18th November 2021 which is impossible.
In addition to the 6 events recorded on the database as ‘gym’ there are two entries under sports participation and described as ‘sports gym’. These occurred at the outset of the pandemic, in February 2020, in Chiba Prefecture, Japan and Nagoya-shi, Japan, consisting of 5 and 10 cases respectively. Further details are not available from the link on the database nor is there any evidence of mitigation measures.
Intergenerational Hong Kong Club Dance Scene accounts for 311 of the 615 cases identified by TAG
Of the 615 cases identified in the report by TAG, over half (311) of them come from a ‘Dance School’ in Hong Kong with secondary transmission linked to bars and restaurants. The source comes from the South China Morning Post and the headline states ‘The dance club scene behind Hong Kong's biggest coronavirus cluster’ The article further states it is an...
'...intergenerational dancing community [where] ‘Eager to impress, well-heeled, typically female members of the community can spend huge sums on younger instructors, even using speedboats to ferry teachers to the city from mainland China for a night on the tiles’,
This suggests it may be linked to night clubs rather than fitness studios or gyms. This should not be compared with well maintained, well organised, socially distanced gym adhering to UK health and safety rules and guidance, but accounts for over 50% of the cases used as evidence of the likelihood of SSE in Welsh facilities. Furthermore, the database also highlights the 311 cases as ‘uncertain’ and therefor it is questionable if Welsh government policy should be based on this evidence.
Reported SSE in Cheonan, South Korean Zumba classes
On February 15th 2020 in Cheonan, South Korea, 27 Zumba instructors attended a 4 hour intense workshop, which resulted in 8 instructors becoming infected and leading to 57 further secondary transmissions to class attendees over the next 24 days. The outbreak was initially reported by the Korean Disease Control and Prevention Agency (KDCPA) which highlighted that 116 cases were investigated in people who had attended classes run by the infected Zumba instructors. However, the route of transmission was directly linked back to the workshop in 65 cases of which 8 were the initial instructors and 57 were students attending subsequent classes. The remaining 51 investigated cases were in-family transmissions (37) or co-workers and acquaintances.
Of the 15 worldwide SSE on the database cited by TAG, 8 SSE and 122 cases can be traced back to the Zumba dance class instructors in Cheonan, South Korea, however half of the cases in the database may have been duplicated. Of the 8 SSE entered in the database, 1 SSE of 65 cases was entered into the database from records by the KDCPA which was referred to as a gym although clearly occurred in a Zumba class, and a further 7 SSE with combined 57 cases were entered from the research by Jang et al., (2020) recorded as Fitness Class. The difference between the two numbers (65 vs 57 cases) could refer to the initial 8 Zumba Class instructors who were infected at the workshop on February 15th 2020 although there is discrepancies in the different sources of data and how they are recorded.
Althouse et al., (2020) state that they ‘consider SSEs to be a single generation of spread where an individual gives rise to a large number of secondary infections’, and the TAG report paraphrases Althouse et al., (2020) for their definition of a SSE in the report. In the Cheonan, South Korea SSE the single generation of spread was at the Zumba dance workshop on 17 February which infected 8 instructors and led to a further 57 secondary transmissions and possibly a further 51. The TAG report uses this outbreak as evidence of the high risk of SSE and justification for prolonged shutting of gyms and states,
“Early in the pandemic, infected fitness instructors in South Korean gyms resulted in 12 SSE within a 24 day period, highlighting the risks posed by vigorous exercise in confined spaces (Jang et al 2020)”.
Jang et al., (2020), does not specifically identify 12 SSE, but highlights that the 8 initially infected Zumba instructors undertook Zumba sessions at 12 sports facilities and the research investigated cases ‘associated with fitness dance classes at 12 sports facilities’. Investigating an association of facilities does not confirm a SSE and at no point in the papers by Jang et al., (2020) or the KDCPA report do they state 12 SSE were identified. Clearly there is discrepancy between the database, research papers by Jang et al., (2020) and KDCPA and the TAG report and suggesting there were 12 SSE is misleading. The main concern is that the TAG report infers the infection was gyms stating ‘infected fitness instructors in South Korean gyms’ when in fact it was Zumba dance class in small studios which caused the event.
The diagram in the KDCPA report identifies the chains of transmission from the Zumba dance class workshop. It indicates that instructors 1, 2 and 3 were so called super spreaders whereas the other instructors spread the virus minimally.
Jang et al., (2020) states the ‘Characteristics that might have led to transmission from the instructors in Cheonan include large class sizes, small spaces, and intensity of the workouts’, and identified that ‘Classes from which secondary COVID-19 cases were identified included 5–22 students in a room ≈60 m2 during 50 minutes of intense exercise’. A room of 8m x 7.5m gives a dimension of 60m2. Under section 4 of the Welsh government guidance on Sport, Leisure and Recreation it is recommended that 100 sq ft or 9.3m2 per person be provided for indoor activities and based on these guidelines only 6 participants (5 students and 1 instructor) would be able to attend a Zumba class in a 60m2 studio, yet up to 22 people were reported to have attended in studios of this size in Cheonan. This provides further evidence of a lack of mitigation measures being implemented. It can only be speculated as to whether the infections were due to the activity (Zumba dance) or the large number of people in a small indoor space, or possibly a combination, although consistently the Welsh government have identified groups of people gathering indoors together in small unventilated spaces as the major cause of spreading of the virus. It should also be noted that Jang et al., (2020) found no cases of transmission from infected instructors when teaching Yoga or Pilates classes. Finally, it should be reemphasised that these cases are linked to Zumba dance classes in small,
oversubscribed dance studios and not open gym settings.
Randomized controlled trial performed in Norway
The TAG report cites a randomized controlled trial conducted by Helsingen et al.,(2020) in Norway which found that out of 1,896 individuals permitted to use gyms only one individual returned a positive test. Randomized controlled trials are regarded as a high level of evidence and the TAG report states,
‘a randomized controlled trial performed in Norway between May 25, 2020 and June 7, 2020 where the incidence of PCR-confirmed SARS-CoV-2 infection in 1,896 individuals permitted to use gyms with a control arm of 1,868 individuals was monitored (Helsingen et al., 2020). Only one individual in the study returned a positive test [which was] due to a 'workplace infection.’
Helsingen et al., (2020) state ‘The positive individual was randomised to training, but had not used the training facility, and the workplace was identified as transmission source’, highlighting that no positive transmissions were recorded relating to the gym.
At the time of the study there was a low incidence rate of 3.5-11 cases per 100,000 population in the region, and therefor inferences about whether the infection rate would be greater if there was higher incidence in the population are difficult, but does suggest that if cases are low risks are low. Of the 1896 individuals who used the gym 38.5% used it on more than 6 days per week which equates to over 4,300 uses in that group alone. The facilities employed mitigating measures including physical distancing of 1 meter for gym floor exercise and 2 meters for high-intensity classes, disinfection and cleaning of each equipment after use by the participant, and enhanced cleaning of facilities and access control (Helsingen et al., 2020). This research strongly suggests that well organised gyms following appropriate mitigation and control measures for the transmission of the virus provides relatively little chances of transmissions.
Sheffield Hallam University confirms ‘extremely’ low levels of Covid-19 risk in gyms
The SafeACTiVE study, commissioned by EuropeActive, reports preliminary findings of a new study from Sheffield Hallam’s Advanced Wellbeing Research Centre (AWRC) and King Juan Carlos University in Spain which analysed more than 62 million fitness facility visits since September and found only 487 positive cases reported from operators based in Germany, France, Sweden, Belgium, Netherlands, Spain, Portugal, Norway, Switzerland, Czech Republic, Poland, Denmark, Luxembourg and the United Kingdom.
‘The findings mitigate public health concerns about the safety of gyms and confirm that fitness clubs are safe environments with relatively low risk of Covid-19 infection. The research provides a strong argument for keeping the fitness and physical activity sector open during any future outbreaks of infectious diseases.’
Section 2, (Mitigate infection through aerosol transmission) of the TAG report Misinterpretation of evidence
To mitigate infection through aerosol transmission Section 2 of the TAG report states,
‘A potential mitigation is the provision of 100% fresh air rather than the recirculation of air within or between indoor spaces, and a rate of 20 litres per second per person is recommended in line with UK Government advice (https://www.gov.uk/guidance/working-safely-during-coronavirus-covid-19/providersof-grassroots-sport-and-gym-leisure-facilities ).’
Ventilation calculations are a specialist area and the statement above is easily misinterpreted, by suggesting that facilities need to provide 20liters per second per person of fresh air into a facility to reduce the transmission through aerosol, which would seem impossible. However, the data comes from the UK gov document ‘Working Safely with Coronavirus’ under ‘Providers of grassroots sport and sports facilities (Section 5 Ventilation of Indoor Sports Facilities) which states,
‘Ventilation is an important part of mitigating against the transmission of COVID-19. Ventilation into the building should be optimised to ensure a fresh air supply is provided to all areas of the facility and increased wherever possible. Particular attention should be given to areas where high intensity exercise takes place.
To achieve this, specific measures should be implemented:
• The maximum occupancy of each indoor facility should be limited by providing a minimum of 100sqft per person. For this figure, the area is the net useable indoor facility space available to members to use, including changing rooms, toilet and wash facilities. Reducing capacity in this way whilst sustaining ventilation flows, will increase the typical current 10l/s/p flow rate of ventilation to at least 20l/s/p, as fewer people are being served by the ventilation system.
• Ventilation systems should provide 100% fresh air and not recirculate air from one space to another.’
The UK guidelines clearly identify that
adequate ventilation is achieved through optimal capacity control namely by providing 100sqft (9.4sq m) per person,
which allows the required ventilation flows to be met.
Section 3: Infection by direct contact with contaminated surfaces and equipment shared between users
The TAG report highlights that equipment may become heavily contaminated by respiratory secretions from infected users and that ‘regular and effective hand hygiene…[and] effective disinfection of equipment between each user represents a primary control’, and therefor appropriate mitigation procedures will hugely reduce risk.
Section 4: Indoor exercise facilities as workplaces and peripheral activities
The report states that ‘Workplace specific guidance should be rigorously adhered to and owners should conduct full COVID secure workplace risk assessments (https://gov.wales/your-responsibilities-employercoronavirus)’, and therefore mitigation procedures are already in place across facilities in Wales. Mitigated by stringent infection prevention measures on the part of instructors and staff and avoiding loud verbal commands by instructors are also recommended.
Limitations
This review is not a full systematic literature review and does not have available data from track and trace in the UK to determine if SSE have occurred in facilities in the UK, but instead focuses on the evidence published in the TAG report. Any new data available about new variants of the virus have not been made available to the public and therefore not included. The author of the report is a gym owner, in Cardiff which could increase bias however the gym fully supports the closure of facilities when transmissions are extremely high such as at alert level four but seeks more robust scientific evidence to be provided should the delayed reopening of gyms be prolonged further as suggested by the First Minister on 19 February 2021. Further investigation, scrutinizing and peer review of this evidence is welcomed and encouraged.
Summary
The evidence provided by the TAG is incomplete, misleading, and has not been fully scrutinised. The overwhelming majority of worldwide SSE recorded on the informal database are not in gyms, (2-4 out of 2007 SSE). There are greater records of cases from dance or group exercise classes in small studios, with large class sizes and no mitigation measures and from a Hong Kong dance club scene. The TAG report claimed 6 of 15 SSE in the worldwide database were from gyms however there is huge errors in the reporting and classification of data and uncertain levels of evidence based on newspaper articles. Of the limited evidence of SSE that may be attributed to gyms, one has 5 uncertain cases reported by a newspaper in Japan and the other 40 cases from a fitness centre in North Carolina, United States which runs 17 group fitness classes per day, with approximately 35 attendees per class, and no evidence of any mitigation procedures.
Of the 615 cases highlighted by the TAG report from the worldwide database as evidence of SSE, 311 cases (51%) are from newspaper reports on an intergenerational Hong Kong club dance scene, 40 cases were duplicated entries from a gym in North Carolina, United States. A further 120 cases came from 2 tightly packed spin studios in Canada and 122 cases can be attributed to a Zumba instructor dance workshop and secondary transmissions in Zumba classes in Cheonan, South Korea where the ratio of participants to space was hugely oversubscribed compared to UK standards and the severity of the pandemic may not have fully been accepted. Crucially there is little evidence of any mitigation or preventative measures to limit transmission employed in any of the cases and vast majority came from dance or fitness classes. The TAG report claimed 12 SSE occurred in Zumba classes from Cheonan, South Korea and whilst 12 facilities were investigated there is no mention of 12 SSE in the two research papers, and a further discrepancy with the corresponding entry in the worldwide database highlighting a need for further investigation.
Strong evidence of the low levels of Covid-19 risk in gyms is provided by a randomized control trial in Norway which found no cases of infection directly linked to gyms in 1,896 gym attendees 40% of whom attended the gym more than 6 times per week and by the extensive research into 62million European gym visits conducted by Sheffield Hallam University confirming ‘extremely’ low levels of Covid-19 risk in gyms’.
Althouse et al., (2020) state
‘Many explosive SSEs have occurred in indoor settings, stoking the pandemic and shaping its spread, such as long-term care facilities, prisons, meat-packing plants, produce processing facilities, fish factories, cruise ships, family gatherings, parties, and nightclubs.’
They do not mention gyms or studios in the research paper on SSE.
This review of the evidence cited by the WAG report, concludes that there is little evidence of SSE in gyms and there should not be prolonged delays in opening providing appropriate mitigation measures and risk assessments are in place identical to those employed in the Autumn. There is greater evidence for the risk of SSE in Dance classes and studios and these activities would best be performed outside in open spaces, online, or in large open buildings where a minimum of 100 sq ft or 9.3m2 per person as recommended by UK and Welsh government can be maintained. Further investigation, scrutiny and peer review of this evidence is welcomed and urgently required as decisions based on the TAG review affect thousands of peoples mental and physical wellbeing, livelihoods and businesses and must be based on rigorous scientific evidence.
Summary Table of Superspreading Events in Gyms, Dance and Fitness Classes identified in the informal database (superspreadingdatabase.com)
References and Links
Althouse, B.M., Wenger, E.A., Miller, J.C., Scarpino, S.V., Allard, A., Hébert-Dufresne, L. and Hu, H., 2020. Superspreading events in the transmission dynamics of SARSCoV-2: Opportunities for interventions and control. PLoS Biology, 18(11), p.e3000897.
Helsingen, L.M., Løberg, M., Refsum, E., Gjøstein, D.K., Wieszczy, P., Olsvik, Ø., Juul, F.E., Barua, I., Jodal, H.C., Herfindal, M. and Mori, Y., 2020. A Randomised Trial of Covid-19 Transmission in Training Facilities. medRxiv. https://doi.org/10.1101/2020.06.24.20138768
Jang, S., Han, S.H. and Rhee, J.Y., 2020. Cluster of coronavirus disease associated with fitness dance classes, South Korea. Emerging infectious diseases, 26(8), p.1917.
Swinkels, K. (2020). SARS-CoV-2 Superspreading Events Around the World [Google Sheet]. Retrieved from www.superspreadingdatabase.com
Further Links
Sheffield Hallam University Research
www.shu.ac.uk/news/all-articles/latest-news/hallam-research-confirms-extremely-low-levels-of-covid-risk-in-gyms
The Korean Disease Control and Prevention Agency report on Zumba Instructors SSE in South Korea is available at
www.cdc.go.kr/board/board.es?mid=a30501000000&bid=0031&list_no=366658&act=view
Or
문서뷰어 (kdca.go.kr)
The Welsh government guidance on Sport, Leisure and Recreation
www.gov.wales/sport-recreation-and-leisure-guidance-phased-return-html#section-58681
The Welsh government Technical Advisory Group (TAG) –‘Environmental Science Subgroup SARS-CoV-2 infection risks at indoor exercise facilities’
https://gov.wales/sites/default/files/publications/2021-01/technical-advisory-group-sars-cov-2-infection-risks-at-indoor-exercise-facilities.pdf
The Worldwide Database of Superspreading events
https://kmswinkels.medium.com/covid-19-superspreading-eventsdatabase-4c0a7aa2342b
UK gov document Working Safely with Coronavirus’. ‘Working Safely with Coronavirus’ under ‘Providers of grassroots sport and sports facilities. Section 5 Ventilation of Indoor Sports Facilities
https://www.gov.uk/guidance/working-safely-during-coronavirus-covid-19/providersof-grassroots-sport-and-gym-leisure-facilities
This review was written by B. Morris. A gym business owner and second year MSc student.
