Influence of ventilation use and occupant behaviour on surface microorganisms in contemporary social housing
Scientific Reports volume 10, Article number: 11841 (2020) Cite this article
In the context of increasingly airtight homes, there is currently little known about the type and diversity of microorganisms in the home, or factors that could affect their abundance, diversity and nature. In this study, we examined the type and prevalence of cultivable microorganisms at eight different sites in 100 homes of older adults located in Glasgow, Scotland. The microbiological sampling was undertaken alongside a household survey that collated information on household demographics, occupant behaviour, building characteristics, antibiotic use and general health information. Each of the sampled sites revealed its own distinct microbiological character, in both species and number of cultivable microbes. While some potential human pathogens were identified, none were found to be multidrug resistant. We examined whether the variability in bacterial communities could be attributed to differences in building characteristics, occupant behaviour or household factors. Sampled sites furnished specific microbiological characteristics which reflected room function and touch frequency. We found that homes that reported opening windows more often were strongly associated with lower numbers of Gram-negative organisms at indoor sites (p < 0.0001). This work offers one of the first detailed analysis of cultivable microbes in homes of older adults and their relationship with building and occupancy related factors, in a UK context.
Homes satisfy our most basic needs for shelter and should be designed to provide a comfortable and safe environment. However, provision of housing is constrained through cost and legislation encompassing appearance, structure, materials, provision of services and energy performance. Most of the time these align to human needs, but there are conflicts, particularly with respect to energy and ventilation and their influence on health. In a bid to reduce energy and carbon emissions, the building sector is delivering increasingly airtight homes that aim to reduce uncontrolled ventilation losses 1 . There are concerns that without improved designed ventilation provision this strategy may lead to a range of unintended consequences including impacts on occupant health 2 . Ventilation affects exposure to a number of elements that are known to influence health, including chemicals, moisture, temperature and microorganisms. There is evidence that poor ventilation may be linked with poor physical and mental health in a number of non-domestic building types 3 , but whilst the literature points to detrimental effects in housing 2 , 4 , this remains seriously under investigated. In particular, there are currently gaps in knowledge about the range and diversity of microorganisms in the domestic environment, particularly in the context of modern airtight homes 5 . People spend a great deal of time in their homes, especially those at the extremes of age, and therefore the indoor microbiome could impact upon human health in ways not yet understood 6 .
Recent research into the real world performance of buildings has begun to reveal significant performance gaps in environmental conditions, especially poor rates of ventilation, particularly in bedrooms 7 . This has led to studies that have examined the consequences of increasing airtightness of modern construction, lack of ventilation, occupant interaction with ventilation, and increasing use of mechanical ventilation 8 , 9 . Recent reviews in the UK by the Royal College of Paediatrics and Child Heath 10 and the National Institute for Health and Care Excellence 11 have undertaken systematic reviews of the literature and recognised the importance of indoor air quality on health and the need to address challenges including with building design. Making buildings resistant to heat loss and draughts has important benefits; higher levels of insulation and airtightness can improve health, for example, through a reduction in cold-related deaths, condensation and mould indoors and reduced fuel poverty 12 , 13 . However, one of the consequences is a separation from the outdoor environment and lower ventilation rates. Theoretical analysis of housing types indicates that potential health consequences of reduced ventilation may include transmission of infectious diseases 14 , and there is emerging evidence that building design affects indoor microflora, with artificial environments created by mechanical ventilation having less diverse microbial communities with a higher presence of pathogens 15 , 16 . Whilst a small number of US studies have demonstrated that architectural design features (such as spatial arrangement or room type) can have an impact on the microbial biogeography of buildings, it remains unclear whether generalizable patterns exist that can be used to inform practice (e.g. through ‘bio-informed’ design) 17 . Moreover, the impact of improved thermal performance (and comfort) standards, energy conservation measures 18 and the creation of hygrothermally stable indoor environments in contemporary housing 8 on indoor microbiology have yet to be fully understood.
There is growing evidence that both building design and human behaviour determine the microbial species present in homes. Care homes have been shown to be a reservoir for antibiotic resistant bacteria including Klebsiella spp. and E. coli 19 , and environmental sampling has shown that high-touch surfaces in home environments may harbour methicillin-resistant Staphylococcus aureus (MRSA) 20 . Multi-drug resistance is having a huge impact in hospitals worldwide but is not commonly investigated in people’s homes; exceptions are the studies by Lax et al., who investigated antimicrobial resistance (AMR) in a hospital setting as well as private homes 21 , 22 . Use of antimicrobial cleaning products have also been highlighted as an emerging AMR risk factor in community settings 23 including use of microbiocidal products used for routine cleaning following confirmed links between disinfectants and resistance 24 . There has even been a call for regulatory use of these products in hospitals 25 ; powerful disinfectants harm the surface ecology, much like antibiotics harm healthy gut flora, permitting naturally tolerant or resistant microflora to survive and create reservoirs of increasingly resistant microbes 26 , 27 .
A number of previous studies have applied DNA sequencing methods to demonstrate the huge diversity in microorganisms present in the built environment 28 . Microbial control and building confinement are known to affect the composition and functional capabilities of the residing microbiome 26 . For example, fungi isolated from home surfaces tend to reflect species found in outside air (where there is no significant moisture damage in the building envelope), while bacterial contamination is more likely to derive from colonised inhabitants and/or their practices 28 ; this suggests that ventilation provision and use may influence the fungal content of the home microbiome. Studies show that the microbial community in a building is influenced by several factors including location and climate 29 , occupant presence and behaviour (including antibiotic use) 30 , presence of pets 28 and ventilation approach 31 . They have also shown that mechanically ventilated buildings have a lower microbial diversity than those that are naturally ventilated, and that the microorganisms present in mechanically ventilated buildings are dominated by human related species, with a much lower presence of environmental species 16 .
Exposure to a wide diversity of microbes has been found to confer protection against certain diseases which has led researchers to suggest that mechanical ventilation may be altering the microbial balance in a building, and could lead to “selection” of microorganisms that are more likely to cause disease or allergies 15 . However, the majority of these studies are conducted outside the UK, with a large proportion in US homes which have different construction, ventilation and climatic conditions. Few studies have performed systematic screening of sites in social housing in order to provide cultivable microorganisms, with a view to establishing the presence of human pathogens and AMR. As drivers for energy reduction have led to buildings becoming more airtight, with reduced ventilation rates and increasing use of mechanical ventilation, insight into the potential consequences of these measures is needed. Given the changes in housing design and construction we need to understand whether our approaches might encourage environmental persistence of a range of pathogens, and evidence from a UK context to support guidance and practice is required.
This study examined methodologies for assessing surface microorganisms in homes and investigated relationships between microorganisms and building characteristics to better understand how they may be affected by the design and use of buildings. The hypothesis is that with reduced ventilation and interaction with the external environment, there will be less diversity in the indoor microbiome and certain organisms may predominate. Whilst ventilation has been identified as a primary driver in the structure of the microbial community in buildings 16 , 31 , 32 , the impact of ventilation type, effectiveness and operation warrants further investigation 33 . In particular, the aim is to examine whether ventilation use leads to a change in the persistence of microorganisms, and to explore both design and lifestyle to assess potential reasons for this.
The study was conducted in two phases, the first of which conducted a household survey and microbial sampling of 100 homes and is reported here. The second undertook more detailed monitoring and analysis of 21 selected homes. We systematically screened specific sites in suburban social housing in order to determine the amount and type of cultivable aerobic bacteria and fungi at key sites in the home. Microbiological data is compared against responses to an extensive household questionnaire containing occupant and housing information. Through this we explore a range of different ventilation and occupancy factors that influence the indoor microbiology of homes and may potentially have an impact on occupant health. We also discuss research methods and protocols that could be applied to larger studies.
Building and occupancy characteristics
The age of the homes ranged from 1995 to 2017, with 34% constructed pre-2010, when building regulations were revised to require airtightness reporting (see Table 1 ). The majority had either one or two bedrooms (94%), with 5% reporting three bedrooms and 1% reporting four or more. Where data was available, air tightness levels ranged from 2.96 to 7.3 m3/h/m2. Airtightness is important in terms of the context for this study as airtight dwellings are entirely reliant on the ventilation provision, and its use by occupants.
Table 1 Building and ventilation characteristics of sampled homes in and around Glasgow.
Full size image
This study presents a detailed analysis of the relationships between cultivable microorganisms and building and occupant parameters in a sample of 100 homes, predominantly occupied by older adults. It is one of a very small number of studies in a UK context and the first to explicitly look for the presence of pathogens and relate this to the building design and use.
Each of the eight sampled sites revealed its own distinct microbiological character, both in the type and number of cultivable microbes. Human pathogens, particularly S.aureus, were more likely to be associated with commonly touched sites such as TV remote, kettle handle and telephone 40 . Whole houses also demonstrated unique microbiological characteristics, with morphologically similar and identifiable microbes observed at multiple sites within the same home 21 . Each home thus displayed its own unique microbiome but with identifiable similarities between other homes according to site.
There was a statistical relationship between homes that opened windows and presence of Gram-negative organisms on sampled sites. This is significant in that it demonstrates a potential effect of window opening on the microbiome and suggests that ventilation design and/or practice may be an important parameter for reducing exposure to specific microorganisms in the home environment. This is likely to be particularly relevant in bedrooms as window opening in these rooms impacts more on long term ventilation rates (i.e. overnight), whereas window opening for thermal comfort will be more short term. It is also possible that bedroom window opening may be an indicator of other health behaviours, for example, occupants who are more health conscious. The presence of both filamentous and yeast-like fungi were not significantly altered by ventilation practices, although the data suggest that fungi were more likely to be found on surfaces if windows are opened infrequently (Fig. 4 ). Environmental Gram-negative organisms are affected by the use of bleach, so it is also possible that cleaning regimens including the use of disinfectants could confound any effect from ventilation practices.
Despite reports of AMR transmission among households, the study offers some grounds for relief, as there was no evidence for multi-drug resistance among recovered isolates that might have pathogenic potential, i.e. S. aureus and human coliforms 39 , 41 . One important reason for this might be the lack of antibiotic pressures in the home as compared with hospitals. Once this pressure is alleviated by patient discharge, home conditions are unlikely to maintain or drive persistence of MDR organisms unless the patient is immunosuppressed or the organism colonises a major carrier site, e.g. MRSA. None of the latter were found among all recovered S.aureus, although it was noted that if one site in the home was positive for S.aureus, it was highly likely to find several other sites contaminated, particularly those that are frequently touched. It is possible that eight sites screened per household were insufficient to isolate the full range of viable microorganisms recoverable. Studies similar to this one in the future should include more surface sites in order to gain a more comprehensive view of the range of recoverable organisms on surfaces in peoples’ homes. It should also be possible to cultivate viral organisms as well, which might offer future recommendations for cleaning the home to minimise transmission of colds and flu among inhabitants.
One of the aims of the study was to develop methodologies for assessing the prevalence of pathogens in housing; this has been rarely undertaken, certainly in the UK. There are a range of practical and ethical issues that arise when gathering data in people’s homes and in this case developing a protocol to gather bacteriological samples alongside survey data was an important outcome. The study methodology was able to gather survey data and microbiological samples from the planned number of sites. The original aim had been to target fewer housing development sites with larger numbers of houses. This would have given greater consistency of location and construction systems; however, these types of development were not available within the timescale of the project. Gaining access to homes, in which there is also accurate constructional information, can be difficult. Approaches were made to larger commercial housing providers for older and retired people, but they were not willing to participate in the research. The study was therefore reliant on using housing associations as gatekeepers which narrows the tenure type. The use of more development sites introduced a larger number of variables and controlling these in studies within housing remains a particular challenge. The study developed a sampling strategy to enable the collection of in-situ samples through the use of dipslides on selected sites, using trained personnel in a commercial survey company. This was a cost-effective method and enabled greater number of surveys in shorter periods of time. However, it was important to ensure that clear protocols were in place to facilitate timely transport of samples to the laboratory. This methodology could be applied to larger studies and, given access to facilities for culturing the samples, could be included as a process on other studies that gather data in homes. It could also be undertaken by healthcare professionals in homes, who could also use the building and occupant survey pro-forma for collection of data about the home. We deliberately used microbiology methods analogous to those used in hospital settings to sample microorganisms, as our primary objective was to determine the presence of potential pathogens in a way that is comparable to healthcare studies. Other studies have made use of sequencing techniques 28 , 29 , 30 , 31 , 32 , 33 which may be a more appropriate methodology where the goal is to characterise the whole of the indoor microbiome. It is recognised that culture-based methods used here have limitations, as they are only able to detect those microorganisms that are both viable and culturable. Sampling efficiency is also a factor that may affect the under reporting of microorganisms, however the dipslides used in this study have been shown to give a comparable recovery to swabbing methods 49 and contact plates 50 .
The survey data is reliant on reported behaviour. Whilst the results are comparable with other similar studies 9 , it should be noted that some differences were noted between reported and actual behaviour, and actual effects of window opening would be dependent on a range of factors such as door opening and external weather conditions. The study was predominantly carried out during the winter months (Nov-April), however the households may have experienced a range of different weather conditions during this period, which may also influence occupant behaviours such as window opening, use of heating and time spent indoors. It is not possible to evaluate the influence of season on the samples collected, but it is acknowledged that this may influence some microbial species, particularly those associated with environmental sources. Previous microbiome studies have shown relationships between outdoor climate conditions and the species found indoors 29 , 31 . The sample used in this study was also small, and on relatively new homes, without obvious defects, or problems such as dampness and mould. A further area of research would therefore be on older existing homes, which may have other environmental and bacteriological characteristics, for example, problems of mould growth may lead to increased ill-health and consequent antibiotic use, potentially increasing the antibiotic pressures in the home environment. While we did not find any significant influence of pet ownership it is possible that pets and indeed family demographics such as children and work patterns of adults may affect behaviours in a way that is not captured within the study. Even within the limitations of the study, the research was able to demonstrate an effect of ventilation on the nature and distribution of bacteria within homes. However, other occupancy and behavioural factors, including cleaning habits clearly influence the presence of bacteria in the home, and larger studies are needed to consider how to separate these effects out.
Overall the study was successful in implementing methods and protocols for the collection of survey data and in particular bacteriological samples within homes in a reasonably efficient and cost-effective manner. The key conclusions from the study are:
The distribution of microorganisms in homes differs between low touch and high touch sites. Low touch locations that are more likely to be contaminated through environmental deposition tend to have higher numbers of microorganisms present and a greater range of different microorganism species.
Homes tended to show consistent characteristics, with specific microorganisms found at one site likely to be found at several other sites. High use of disinfectants appeared to reduce the diversity of microorganisms found within a home, and both smoking and recent antibiotic use were shown to be correlated to a reduced presence of bacteria.
Ventilation provision and use has an impact on the presence of Gram-negative bacteria, with increased window opening reducing the likelihood of finding Gram-negative isolates. There is some indication that reduced ventilation also reduced the microbiological diversity, and in the context of a shift to mechanical ventilation, this is of further interest. Greater evaluation of the hygrothermal conditions of homes that contribute to environments that may support pathogens (for example warmer, wetter homes) is also of further interest.
The data presented here identified a number of microorganisms that could be pathogenic, including Klebsiella pneumoniae and Enterobacter cloacae, however we found no evidence for home contamination of multidrug resistant pathogens. This may perhaps inform policy about the relative benefits of the home environment in terms of exposure to bacteria, which may be relevant to processes for hospital discharge. However, in this study the homes were selected from a constructional perspective, and recent antibiotic consumption was low. An alternative approach would be to identify households through a clinical route, for example patients being prescribed antibiotics or with chronic health conditions to evaluate conditions in homes where antibiotic use is more prevalent.
Recruitment of homes
The study was conducted between November 2017 and April 2018. Households were recruited from seventeen developments across Glasgow, Livingston and Ayr in Scotland UK, covering both urban and rural areas (Fig. 7 ). Of the 312 households that were initially approached, 109 participated in the occupant survey and 100 agreed to microbial sampling.
Full size image
Dipslides were pressed onto chosen sites (if present) for 5–10 s at a pressure of approximately 25 g/cm2 without overlap between sampled areas 48 . The slides were replaced in sterile containers and transported to the microbiology laboratory on the day of collection. After loosening caps, the dipslides were incubated for 48 h in air at 35 °C before processing. Sampling was performed in accordance with recognised practices from the Food Standards Agency. Bacteria and fungi were quantified for each site by assessing growth on nutrient agar according to manufacturer’s instructions. Growth on nutrient agar supplied total aerobic colony counts (ACC) per cm2 which were classified as follows: no growth (NG) 0 cfu/cm2; scanty growth (SG) 2.5 cfu/cm2; light growth (LG) 12 cfu/cm2; moderate growth (MG) 40 cfu/cm2; heavy growth (HG) 100 cfu/cm2); and very heavy growth (VGH) 250 cfu/cm2. This is comparable to approaches used previously in hospital sampling studies 46 , 47 , 48 . Selective agar highlighted potential coagulase-positive staphylococci, which were sub-cultured onto Staphylococcus aureus Identification (SAID) agar (Oxoid Ltd, UK), followed by automated susceptibility testing (VITEK2™) according to routine laboratory protocol. The reader also noted colonial types, morphology and fungi on nutrient agar and performed Gram-stains on a maximum of four cfus per slide, thought to indicate Gram-negative species. Those confirmed as Gram-negative bacilli were screened on UTI selective agar, plated out for purity and identified and characterised by VITEK2, including antimicrobial susceptibility testing.
All methods were carried out in accordance with relevant guidelines and recommendations. The microbial analysis was performed in a CPA accredited clinical laboratory, in accordance with recommendations and standard practices from the Institute of Biomedical Sciences and the Royal College of Pathologists.
Data from the microbial sampling was combined with occupant survey data for each house. Quantitative analysis of microbial data was carried out using the contamination values indicated above, enabling a mean concentration for ACC across all sites in each house to be calculated. We recorded presence/absence of seven categories of microorganisms identified on the nutrient agar samples: Staphylococcus spp.; Micrococcus spp.; Bacillus spp.; filamentous fungi and yeasts; other Gram positive cocci and rods; Gram negative rods; Gram negative cocci. A diversity measure was calculated to indicate the proportion of these categories present at each site; a diversity of 1 would indicate that all 7 species were present, a diversity of 0 would indicate none.
Household survey data was converted into numerical responses. Questions with a yes–no answer were allocated a value of 1 or 0 respectively, questions with more than one response were given a number for the category. Some additional values were calculated for the analysis based on the survey responses. The questionnaire asked participants how often windows were usually opened in the home during the day and at night, throughout the winter season. Participants responded using a 5-point Likert scale, from ‘Never’ to ‘All the time’ for specific rooms including the kitchen, living room, bedroom(s) and bathroom(s). A total window opening frequency value (%) for winter was calculated by assigning scores to the ordinal data and converting to interval data, using the following weighting: no window/never = 0, monthly = 1, weekly = 2, daily = 3, all the time = 4. A whole house percentage was calculated based on the number of rooms. The value represents the weighted frequency of window opening in the home in winter, during the day and at night. For instance, a value of 0% indicates that windows were reportedly never opened (or no window was present) in all rooms (living room, kitchen, bathroom and bedroom(s)) during the day or night, with 100% indicating all windows were reportedly opened all the time.
The survey asked for information about eight common disinfectant products as well as an additional question that asked about other products used. Almost all of the homes indicated that they had used a disinfectant product or bleach in the last week. To capture the level of disinfectant use, a numerical average was taken of the responses to disinfectant products; a value of 1 indicates the household used 8 different products, a value of 0 indicates none. An additional variable was constructed to indicate whether bleach was used as a yes–no response.
All statistical analysis was carried out using R software (version 4.3). Shapiro–Wilk test was used to examine non-normally distributed microbial counts for ACC on both nutrient agar (p = 2.94E−10) and the selective agar (p = 3.23E−12). Log10 transformation returns normally distributed variables (p = 0.44 and p = 0.137 respectively) and hence was used in linear regression analysis. Welch’s t-test was used to compare between means with unequal group variances, and ANOVA enabled assessment of the difference between sample sites. Where appropriate the Kruskal–Wallis test is used normality is not upheld, sample sizes were small. Post-hoc testing using Tukey HSD after ANOVA or Dunn’s test with Holm-Sidak correction after Kruskal–Wallis allowed multiple site comparison where appropriate.
Data is available at https://doi.org/10.5518/804 .
Ministry of Housing, Communities & Local Government, The Future Homes Standard, 2019 Consultation on changes to Part L (conservation of fuel and power) and Part F (ventilation) of the Building Regulations for new dwellings, https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/852605/Future_Homes_Standard_2019_Consultation.pdf (2019).
Shrubsole, C., Macmillan, A., Davies, M. & May, N. 100 unintended consequences of policies to improve the energy efficiency of the UK housing stock. Indoor Built Environ. 23, 340–352 (2014).
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information 2.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .
About this article
Cite this article
Sharpe, T., McGill, G., Dancer, S.J. et al. Influence of ventilation use and occupant behaviour on surface microorganisms in contemporary social housing. Sci Rep 10, 11841 (2020). https://doi.org/10.1038/s41598-020-68809-2