Defining a Healthy Microbiome of the Built Environment

Abstract

Built structures host complex and dynamic microbial communities whose assembly is shaped by occupancy, ventilation, building materials, cleaning practices, connections to outdoor environments, and other factors. These communities, collectively described as the microbiome of the built environment (MoBE), are increasingly recognized as critical to human health, yet there is currently no consensus definition of what constitutes a healthy MoBE. This gap impedes the development of engineering efforts to balance microbial diversity and immune support with the minimization of infectious disease risks. We propose three pillars that together define a healthy MoBE: (i) pathogens below context-specific risk thresholds; (ii) non-sterility, as sterility is likely to be suboptimal for human health in most contexts; and (iii) resilient diversity, as more diverse communities may resist pathogen colonization and recover more rapidly from perturbations. We synthesize evidence for each pillar, propose a framework for feedback-controlled MoBE engineering, and discuss how pillar priorities should be weighted across building types and occupant populations. We identify critical research priorities to advance actionable performance standards in support of healthy buildings.

Introduction

Humans spend an estimated 90% of their lives indoors. The built environment (BE) – encompassing homes, workplaces, schools, and healthcare settings – therefore constitutes the principal arena of human-microbe interaction. Decades of research have documented that indoor spaces harbor complex and dynamic microbial communities shaped by a range of variables including ventilation, occupancy, materials, and cleaning practices. These communities are a function of occupants’ microbiomes and that of the wider environment. In turn, they modulate occupant health via pathogen transmission and infection, immunological effects, the spread of antimicrobial resistance (AMR) genes, and other interactions.

Yet a fundamental challenge persists: there is no universal, consensus definition of a “healthy” microbiome of the built environment (MoBE). Without such a definition, engineering the MoBE to achieve health-related outcomes is like designing a water treatment system without a water quality standard: how would we know if we achieved the desired result? Clearly, we need measurable targets and verifiable outcomes. Outcome-based standards like ASHRAE 62.1 (Ventilation for Acceptable Indoor Air Quality) and ASHRAE Standard 241 (Control of Infectious Aerosols) rely on performance metrics intended to assess microbial control measures; these standards are driving important innovations in building performance, including to achieve health outcomes.

In this Perspective, we propose a broad working definition of a healthy MoBE built on three empirically grounded and measurable pillars: (i) risk-minimized pathogens, (ii) non-sterility, and (iii) resilient diversity. We describe existing evidence supporting each pillar and propose priorities to advance the framework – including indicators, targets, and interventions – that can serve as a useful basis for performance-based MoBE engineering.

Recent reviews have substantially advanced the conceptual landscape of indoor microbiome research. Gilbert and Hartmann comprehensively reviewed the indoor microbiome and human health, and Bosch et al. articulated the potential importance of the built-environment microbiome within frameworks of multispecies health and health equity. In this Perspective, we complement and extend these contributions by focusing specifically on engineering operationalization. Building on an existing rich literature describing the ecology of the MoBE and emerging epidemiological evidence, we focus on the question of how to define a healthy MoBE in actionable terms for feedback-controlled building engineering and eventual translation into building performance standards. We further build on insights from quantitative microbial risk assessment (QMRA) and systems engineering theory to offer a conceptual foundation for health-based building standards that can be further refined in empirical studies as the field matures.

The Case for a Definition

The MoBE is variously defined but includes bacteria, archaea, viruses, protozoa, fungi, and other microscopic eukarya present in or on surfaces, air, water, and other media in human-constructed spaces. Indoor microbial communities originate in multiple sources: building occupants themselves (e.g., skin, respiratory tract, gut, exhalations); synanthropic or domesticated animals; mechanical transport of microbes into spaces on feet or objects; outdoor air entering via ventilation; plumbing and HVAC systems; and building materials. Unlike natural ecosystems, the BE includes anthropogenic factors that can strongly influence the distribution, assembly, succession, and other characteristics of indoor microbial communities: some of these include temperature and humidity control, ventilation, plumbing characteristics, architectural design features, building uses, natural or artificial lighting, chemical and other disinfectants or microbial controls, density of occupation, and control of animals and pests. As a result, the MoBE is generally less diverse than outdoor or soil communities but consistently includes human-associated taxa and is characterized in part by selective pressures of modern sanitation including disinfection and other environmental controls.

Defining a “healthy” microbiome in any context is notoriously difficult. For the human gut, the field has largely abandoned the search for a single universal definition of microbiome health, recognizing instead that diversity and functional redundancy may be more important than specific taxonomic profiles. The human gut is characterized by a very narrow temperature range within a complex but constrained biological system, but still exhibits enormous diversity and variability within and between individuals. The BE exhibits even more extreme variability in environmental conditions and other factors influencing the MoBE. BE microbiomes are also typically low-biomass and subject to high spatial and temporal variability, posing real barriers to representative measurement and engineering controls. Like the human gut, there is unlikely to ever be a specific “healthy” definition of the MoBE that includes specific taxa or a universally optimal distribution of specific microbial groups or functions; the inherent diversity and dynamics of these systems precludes it.

Critically, the framework we propose does not envision the MoBE as a static system, to be held at a fixed optimum target state. Indoor microbial communities are better understood as systems in continuous ecological succession, shaped daily and seasonally by variations in building occupants, ventilation, cleaning, and many other drivers. Such a complex system may be steered to achieve desired states consistent with our understanding of “healthy”, but generally cannot be precisely controlled. Rather, the goal is to achieve and maintain microbial communities within a defined envelope of acceptable states.

A further complexity is the bidirectional relationship between building occupants and the MoBE, particularly given that the whole point of a “healthy MoBE” is to support the health of building occupants. In most contexts, the people inhabiting buildings represent dominant sources of indoor microbial communities via skin shedding, respiratory emissions, and other routes, and their microbiome states vary considerably as a function of age, health status, diet, antibiotic use, and other factors. But the MoBE also influences occupant microbiomes and immune states in turn, creating occupant–building feedback loops that must be accounted for. Engineering buildings without accounting for this bidirectionality would be unrealistically simplistic. Spatial heterogeneity within buildings adds further complexity. Floors, HVAC, plumbing biofilms, and occupied zones each represent distinct microhabitats with characteristic community compositions, different disturbance regimes, and different roles in the overall building microbiome. These things are in turn functions of the building’s design and use, which are key elements of architecture. Exploring and incorporating these sources of complexity and feedback are important priorities for future elaboration of the framework we propose.

Despite inherent complexity, a pragmatic engineering definition needs only to be measurable, adaptable, and consistent with our best (admittedly, evolving) understanding of what healthy MoBEs look like, in terms of empirical or predicted causal relationships to specific beneficial outcomes as engineering endpoints. Chiefly, that outcome is the health and well-being of building occupants. To this end, any definition of a healthy MoBE should be (a) actionable, in terms of translation into measurable indicators and performance targets; (b) context-sensitive, recognizing that buildings are different and use cases matter; and (c) theoretically grounded, informed by an understanding of the mechanisms by which the MoBE affects the ends we want to achieve.

Based on this approach, we propose three pillars supporting our definition for a healthy MoBE.

Pillar I: Pathogens Below Tolerable Risk

The clearest connection between the MoBE and human health is via exposures to pathogenic microbes, long a primary focus of environmental health microbiology. Measuring pathogenic microbes and estimating their potential harms – and intervention strategies to interrupt transmission – has been foundational to the development of the BE as a whole, including plumbing and ventilation systems. The effective control of pathogens is still a central priority for any MoBE strategy, even in low-burden populations in high-income countries. For example, healthcare associated infections affect approximately 1 in 31 hospitalized patients in the United States, costing an estimated $28–45 billion annually. HAIs overall contribute substantially global morbidity and mortality. The ESKAPEE pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, Enterobacter spp., and Escherichia coli) represent the principal AMR threats in hospital BEs, where they persist on surfaces and in plumbing for extended periods. In all buildings, respiratory pathogens – including influenza, SARS-CoV-2, and Mycobacterium tuberculosis – may be transmitted via airborne and fomite routes shaped by ventilation, occupancy, and other building dynamics. Mold presents persistent if poorly characterized health risks in many settings. Emerging and reemerging pathogens like Candida auris and the measles virus are good reminders that infectious diseases are unlikely to be wholly excluded from BEs any time soon.

A foundational insight in environmental microbiology is that, in most contexts, pathogens can never be wholly eliminated, only reduced to some tolerable level. We therefore propose that a healthy MoBE maintains known pathogens, including opportunistic pathogens, below context-specific risk thresholds. Here, we build on quantitative microbial risk assessment (QMRA), a framework that makes use of pathogen dose-response relationships and a priori tolerable risk thresholds to derive quantitative performance targets for microbial control strategies or acceptable microbial hazards in environmental media. These methods have been used most prominently in the context of waterborne infection and disease risks, but are now commonly applied to other pathways of exposure including via fomites and air. For a given building type and occupant population, QMRA links pathogen concentration (e.g., genome copies per m³ air or per cm² surface) to infection probability, enabling derivation of risk-informed thresholds.

Target risk thresholds should be context-specific. A tolerable MRSA surface density in an outpatient waiting room may differ from what would be acceptable in a bone marrow transplant unit, or in an office building, because the dose-response relationships between the bacterium and health outcomes would be different in the populations of people inhabiting these spaces. The stakes are higher for microbial control measures among immunocompromised individuals. Risks are therefore always context-specific, and so too must be our approaches to managing them. ASHRAE Standard 241⁷ operationalizes an analogous risk-based approach for aerosols, specifying equivalent clean airflow rates (ECAi) that reduce inhalation exposure below defined infection probability targets. Methodological convergence between different approaches to taking microbial (or microbe-associated but non-microbial) data and predicting risk is a major next step, combined with establishing tolerable risk thresholds specific to building types or use cases. The actual levels of tolerable risk vary by application and framework, but common standards are 1 × 10^–6 disability-adjusted life years (DALYs) per person per year, used by the World Health Organization, or 1 in 10,000 annual risk of infection per person used by the US EPA for infectious disease risks. The latter may be broadly applicable across buildings with healthy occupants and could be considered as an initial target for many use cases.

The risk-minimized pathogen pillar should also include consideration of opportunistic pathogens and the environmental resistome. Many organisms that rarely cause infection in healthy adults, such as Pseudomonas aeruginosa, Stenotrophomonas maltophilia, Aspergillus fumigatus, Candida auris, and non-tuberculous mycobacteria (NTM), pose risks to immunocompromised individuals, the elderly, and infants. AMR enrichment in the MoBE should be treated as a risk endpoint in its own right. Aggressively disinfected environments may exhibit not only reduced microbial diversity but also higher diversity in resistance genes, with a shift toward Gram-negative Proteobacteria enriched for AMR determinants. A building that meets risk-based pathogen density targets but harbors an enriched resistome may still pose non-negligible public health hazards, particularly in healthcare settings where multi-drug-resistant organisms are known threats. Pillar 1 therefore considers not only pathogen abundance below QMRA-derived thresholds of tolerable risk, but also surveillance and minimization of AMR enrichment and the selective conditions that promote AMR emergence and dissemination.

Pillar II: Non-Sterility

Conventional infection control is predicated on reducing total microbial load through disinfection, with sterility as a normative goal. This approach is both effective and necessary in high-risk applications, like in sterilizing surgical instruments or in food packaging. Sterility is neither achievable nor desirable as a universal indoor standard, however. Disinfectants and other microbial control measures never eliminate all microbes, and some groups like spore-forming bacteria (Clostridioides difficile, Bacillus spp.), mycobacteria, fungi, or protozoan oocysts are especially resistant to treatment and may persist for long periods. Critically, aggressive disinfection creates ecological vacuums or “microbial deserts” that are susceptible to rapid recolonization by tolerant organisms with potential priority-effect advantages over other microbial communities. Because recolonization and emergent resistance to treatment are observable core features of microbial ecosystem responses in the BE, we propose that non-sterility, i.e., the deliberate avoidance of broadly sterile conditions, is a defining characteristic of a healthy MoBE.

This pillar suggests a need to reduce broad-spectrum microbial disinfection in many settings in favor of more microbiome-friendly cleaning methods. Examples include so-called probiotic-based sanitation (PBS) systems, in which presumptively benign or “probiotic” microbes (often spore-forming Bacillus strains) are applied inside buildings to competitively displace pathogens as an alternative to broad-spectrum or indiscriminate disinfection. Hospital-based trials of PBS have demonstrated reductions in HAIs and pathogens on surfaces. The evidence for PBS is promising but preliminary, however. Long-term consequences of routine probiotic inoculation, including potential shifts in the resistome or selection for environmentally adapted Bacillus variants, have not been systematically evaluated. Introduction of other specific microbial taxa have not been extensively studied as a means of altering the MoBE; this is a promising avenue for research.

Beyond PBS and similar approaches, other strategies hold potential for avoiding the creation of microbial deserts in buildings, either as standalone interventions or as complements to more conventional microbial management. Ventilation strategies that increase outdoor air exchange can introduce outdoor microbial diversity that can in turn diversify and potentially stabilize indoor microbial communities, shifting them away from pathogen-enriched states. − Surface materials that effectively resist biofilm formation without broad-spectrum biocidal activity could be selected over those that exert selective pressure for emergent resistance. Novel microbiome-friendly cleaning solutions are increasingly common, with an emerging evidence base. Following cleaning with probiotic seeding to prevent recolonization or the introduction of outside air may be important complements to conventional pathogen inactivation strategies. The goal is to leave no microbial deserts behind, whatever the method.

In addition to its other potential benefits, non-sterility in the BE is associated with appropriate immune development and training. The hygiene hypothesis and its successor, the “old friends” hypothesis, posit that reduced microbial exposure in the modern BE contributes to the high and rising prevalence of allergic and autoimmune conditions in high-income countries. While these causal relationships remain incompletely characterized, epidemiological evidence from high-income settings consistently shows associations between higher microbial diversity in the BE and some reduction in chronic disease risk in the people exposed to that diversity. A non-sterile MoBE that results in diverse, mostly nonpathogenic microbial exposures is therefore consistent with both infection prevention and long-term immune health objectives.

Pillar III: Resilient Diversity

Diverse microbial communities are more stable, more resistant to invasion, and more capable of recovering from disturbance. Genetic and functional diversity convey resilience via redundancy, the ability of a microbial ecosystem to undertake similar functions or to occupy similar specific niches because multiple microbial populations share ecological or metabolic roles. Resilience determines whether a MoBE can maintain a healthy state in the face of perturbations including cleaning events, occupancy or building operational changes, seasonal variation, and pathogen introduction or colonization.

In the BE, diversity operates at multiple levels and can be measured in different ways. Alpha diversity (richness and evenness within a space) determines the functional redundancy available to absorb disturbance. Beta diversity (variation across spaces or time points) reflects the adaptive capacity of the system to reestablish community composition after perturbation. Functional diversity may be more directly relevant to MoBE resilience than taxonomic diversity per se, covering redundancy of important metabolic pathways, mechanisms for competitive exclusion, nutrient cycling functions, and horizontal gene transfer dynamics across microbial communities. Two buildings with identical alpha diversity indices but substantially different functional gene content could have very different capacities to resist pathogen colonization. Given the rapid development of functional metagenomics, coverage of key metabolic pathways or functional gene categories is increasingly measurable. Such metrics could incorporate taxonomic diversity in MoBE assessments. We therefore propose that Pillar 3’s goal of diversity be understood as encompassing both taxonomic and functional resilience.

Resistance to pathogen introduction and colonization is a key reason why diversity should be a feature of any health MoBE. Building on classic insights into resilience in ecological systems, Van Elsas et al. (2012) demonstrated a strong correlation between species richness in soils and resistance to pathogenic E. coli invasion, a finding that has been reproduced across ecosystems types and is known as the “diversity-invasion effect”. It is therefore possible that more diverse MoBEs will be more resistant to colonization when pathogens are introduced via infected occupants, contaminated objects or surfaces, outdoor air, or other sources. Colonization resistance has been extensively described in the context of the gut microbiome as well and is a mechanistic analog for MoBE resilience through diversity. Existing diverse taxa compete for niches, produce antimicrobial compounds, and modulate nutrient availability in ways that can prevent or slow down pathogen establishment. Diverse microbial communities also buffer against horizontal gene transfer of AMR determinants. Competitive occupancy of available niches limits the selective advantage accruing to AMR-enriched colonizers following disturbance, and diverse communities are associated with lower overall resistome enrichment compared to disinfection-disrupted communities.

Resilience is quantifiable through longitudinal monitoring before and following interventions. In the gut microbiome, antibiotic perturbation studies have shown that recovery trajectories are diverse: some microbial communities return to pre-disturbance states within weeks; others converge to altered stable states with impaired function. We expect similar dynamics in the MoBE, for example in response to broad-spectrum or targeted disinfection. Resilient MoBEs would show rapid recovery following microbial control measures. Conversely, fragile MoBEs might exhibit extended dysbiosis, creating conditions that may be ripe for pathogen invasion and colonization. We note, however, that the dynamics of colonization resistance have not been comprehensively interrogated in the highly complex, diverse, low-biomass systems typical of MoBEs, which differ in important ways from gut and soil systems that have been better studied. Further empirical work and synthesis of findings across studies to reveal insights into microbial invasion and resistance in the context of MoBEs are important priorities for further research.

Diversity can be quantified as part of any definition of a healthy MoBE, but, again, context matters because the type of diversity and how it is measured may tell us different things about the relative health of buildings. Alpha diversity indices (e.g., Shannon, Simpson, Faith’s PD) can be calculated from 16S rRNA amplicon sequencing or shotgun metagenomics and compared longitudinally. To develop a standard metric, diversity measure targets can be validated against reference communities from well-characterized healthy buildings, or derived from empirical studies relating indoor diversity and measured health outcomes. Functional diversity metrics like coverage of key metabolic pathways, resistome diversity, or presence of keystone taxa could complement taxonomic diversity metrics and may prove more directly relevant to some health outcomes. The resistome in particular serves as a cross-cutting indicator spanning all three pillars: resistome diversity and HGT potential are properties of community composition (Pillars 2 and 3) with direct health implications (Pillar 1).

It is important to note that the non-sterility and resilient diversity pillars are framed primarily within high-income country contexts where the endemic burden of infectious disease is much reduced from historical highs. The hygiene hypothesis, studies suggesting benefits associated with exposure to farms, and biodiversity intervention trials that support Pillar 2 come overwhelmingly from high-income settings, where pathogenic exposures are generally much lower than in high-burden environments. In contexts where Pillar 1 remains the priority – where diarrheal disease, acute respiratory infection, or endemic illness still impose major mortality and morbidity burdens borne disproportionately by the world’s poorest people – the non-sterility and diversity arguments are necessarily subordinate to pathogen control. A global perspective on healthy MoBE definition must therefore acknowledge that the relative priority of the three pillars varies with epidemiological context, and this Perspective is most directly applicable to low-burden, high-income settings. We intend the framework to be adapted for other contexts as evidence accumulates and as Pillar 1 targets are achieved globally, however, and that must require a balancing of these priorities as the context is considered.

Engineering the MoBE

The three-pillar definition we propose is designed to enable MoBE engineering via feedback-control and performance-based building standards. We envision integrating building microbial measurement, interpretation, and modification into the BE by coupling microbial assays with QMRA-enabled interpretation of risk and targeted intervention strategies that reduce pathogens below a priori tolerable risk thresholds (Pillar 1) while not creating sterile environments (Pillar 2) and prioritizing diversity (Pillar 3). Conceptually, we propose a straightforward feedback-control system: measurement of the community state, comparison with defined quantitative targets defining “healthy”, and implementation of interventions to alter the system. A thermostat offers a conceptual, if simplistic analogy: temperature is measured, compared with a set point, and heating or cooling is applied to reduce the deviation. The application of such a framework to MoBE engineering is far more complex, commensurate with the complexity and diversity of MoBE systems generally.

The gap between this conceptual model and current technical capabilities defines the core engineering research agenda for MoBE management: developing and carefully validating rapid, proxy-based sensors validated for low-biomass environments; defining quantitative performance targets that define “healthy”; and understanding intervention effects in real-world systems are all needed. Relationships between interventions and community states are nonlinear, history-dependent, and context-specific. Some interventions, such as disinfection practices that select for AMR-enriched communities, are not readily reversible.

A practical question follows from the coexistence of all three pillars: how should we navigate settings where aggressive disinfection is necessary for pathogen control (Pillar 1) but where doing so simultaneously erodes the resident microbial community (Pillars 2 and 3)? We suggest that context governs which priorities take precedence, and that the appropriate hierarchy shifts across a spectrum of building types and occupant vulnerability. At one end, intensive care units and bone marrow transplant facilities represent settings where immunocompromised patients face life-threatening infection risks: Pillar 1 rightly dominates, and the tolerance for non-sterility is necessarily low. At the other end, primary schools and residential environments represent settings where the evidence supports substantially greater weight for Pillars 2 and 3, with Pillar 1 approaching microbial controls less aggressively, appropriate for generally healthy occupants. Between these use cases, the appropriate emphasis would depend on occupant vulnerability, endemic pathogen pressure, and specific building functions. Further developing an explicit heuristic to guide MoBE engineering approaches that are appropriate for the context is an important next step for operationalizing the framework.

Research Priorities

Engineering healthy MoBEs will require serious investment in research. Below, we outline these needs in approximate order of logical dependency, recognizing that some investments are prerequisites for others and that they all overlap in important ways. Coordination across groups working in MoBE engineering will help accelerate progress.

Measurement

The feedback-control paradigm we propose requires reducing time-to-result so that action can be taken when and where microbial exposures present unacceptable risks. Measurement is therefore the first dependency in the research chain: advances in QMRA application, diversity benchmarking, and epidemiological validation all require better measurement methods as their foundation. Current methods for monitoring the MoBE are typically too slow to be straightforwardly actionable, and therefore our understanding of what is happening in MoBE systems tends to be retrospective. We encourage the development of rapid, targeted measurement methods and biosensors, acknowledging that such measures can realistically only be proxies or indicators for low-biomass microbial communities in most cases. Proposed indicators would need to be validated in terms of their predictive value for specific microbes of interest. Current sequencing methods used in describing the MoBE broadly are not typically quantitative, though important methodological advances are helping to overcome this limitation. QMRA requires microbial data that is both quantitative and captures viability, typically at much lower limits of detection than is usual in common sequencing and bioinformatics pipelines or in biosensors. A further challenge specific to the MoBE is the low-biomass nature of many indoor environments: diversity data from such samples are sensitive to contamination, sequencing depth, and extraction efficiency, all of which can substantially alter apparent community composition, and any measurement framework must account for these limitations. Establishing platform-specific validation standards and standardized sampling protocols are long-term projects in the field, allowing for diversity benchmarks to be meaningfully compared across studies or applied in building standards.

While the MoBE definitionally includes viruses, and much of the public health case for engineering the MoBE is based on what we know about the control of respiratory viruses, our understanding of microbial communities inside buildings is mainly limited to bacteria. Viruses and other non-bacterial constituents of the MoBE are not typically captured in the sequencing approaches that have been most commonly use to characterize these communities. Bacteriophages represent a particularly underexplored component. Phages shape bacterial community dynamics, and phage-mediated horizontal gene transfer is a key mechanism for AMR gene dissemination, including among pathogens of interest in MoBE systems. Measurement approaches to the MoBE should seek to include the indoor virome in developing a fuller picture of microbial community dynamics. Doing so would help advance our understanding of the roles phages play in community stability, resistome dynamics, and potential phage-based interventions; the role of mycoviruses in influencing fungal molds; and how buildings can operate most effectively to control human viruses.

Advancing QMRA for Applications Across the BE

QMRA frameworks are well established for drinking water and food safety but remain immature for indoor air and surface environments, where estimating microbial doses tends to be less straightforward. For example, for a given pathogen in dust on the floor or on furniture, estimating the number of viable microbes ingested or inhaled is subject to a great deal of uncertainty. Extending QMRA to include additional bacterial pathogens, fungi, and AMR organisms, and mainstreaming the use of molecular and other data in QMRA, would advance the field substantially. AI-coupled real-time analyses or models incorporating and propagating uncertainty could also help.

QMRA is based on dose-response relationships, which have heretofore been developed exclusively for pathogens. Given the emerging evidence around beneficial microbial exposures, however, we encourage the consideration and development of positive dose-response relationships that could also be used in assessing the potential beneficial impacts of microbes in the BE on human health. Though there are few certain causal relationships here, QMRA does provide a quantitative framework to speculatively model positive health outcomes as a function of microbial exposures. Evidence from studies in mice have suggested potential protective effects against the development of allergies associated with exposures to Acinetobacter lwoffii and Lactococcus lactis, for example, which have been supposed to be related to the so-called “farm effect” on child health and immunological development observed in several large observational studies in Europe. Exposure to diverse microbiomes may have immune-modulating effects, broadly, pointing to the potential to derive positive dose-response relationships between diversity indices and beneficial outcomes. Quantifying these relationships would be key to operationalizing them in an engineering context.

Epidemiological Validation

A clear, fundamental, and widely appreciated current gap in evidence is the lack of unambiguous, empirical epidemiological studies that show a causal relationship between the MoBE and health outcomes outside of known pathogen exposures in isolation. Well-designed randomized controlled trials and other experimental and observational studies are needed that capture key aspects of any definition of a healthy MoBE – such as the three pillar model we suggest, or otherwise defined – and its effects on measurable health outcomes. Such trials would require longitudinal, multi-omics microbiome profiling of buildings concurrent with health outcome surveillance, across a range of building types and populations. Because effects from trials may be very context-specific and therefore may have limited external validity, many such trials would be needed to fully capture MoBE effects.

Diversity and Resilience Benchmarks

While the relationship between diversity and pathogen invasion resistance is well established microbial ecology, quantitative benchmarks for indoor community diversity, below which resilience is expected to fail, do not yet exist. Establishing these benchmarks will require longitudinal studies that track community diversity and stability across disturbance events (e.g., cleaning cycles, occupancy changes, introduction of pathogens), linking microbial community states to health outcomes. Experimental trials introducing defined levels of microbial diversity (e.g., via ventilation modification or inoculation) and monitoring subsequent pathogen colonization could be helpful. Such benchmarks will also require platform-specific validation and standardized sampling protocols to ensure that diversity measurements are comparable across studies, building types, and geographic settings.

Interventions

The history of environmental health microbiology, from cholera to SARS-CoV-2, has been predominantly oriented toward interventions that were intended to be as effective as possible in completely eliminating microbial hazards. These strategies have been enormously important in creating the modern world by reducing disease, suffering, and death. They still have those effects today anywhere known pathogens are circulating, and we discount that progress at our peril: we are perennially reminded that infectious disease remains a primary threat to human health and well-being, and erring on the side of creating more hygienic environments is well-justified on public health grounds. Water supply, water treatment, sanitation, hygiene, antibiotics, improved air quality, food hygiene, and other measures are foundational to human prosperity. The work continues to extend their benefits to everyone.

While highly effective, broad-spectrum microbial control methods will still have an important role to play in the BE, our success in reducing endemic and epidemic diseases via environmental controls has had some adverse consequences. The rise of treatment-resistant pathogens, the possibility of worsening chronic conditions, and the creation of environments that are paradoxically more pathogen-enriched as a result of too-effective cleaning are distinct challenges we now face. Novel interventions that work with the MoBE rather than against it have important potential as we evolve our microbial control strategies based on our more nuanced understanding of the complexity of microbial ecosystems in the BE and host-environment interactions. These certainly include important ideas described previously and may call for changes in materials; building operation (e.g., to enhance ventilation); targeted disinfection (pathogen-specific protocols that avoid broad-spectrum microbiocidal activity); probiotic-based sanitation that may include competitive exclusion of pathogens via Bacillus or other bacterial and fungal-based cleaning products; engineered microbial inoculants, including those harnessing synthetic biology; targeted uses of bacteriophages and mycoviruses; physical changes to plumbing and fixtures; engineered surfaces; and many other methods, either as standalone approaches, in combination, or in concert with more traditional interventions to control microbes. There is a large and growing community of researchers and innovators working in this space, with a wealth of ideas on how to apply new approaches to engineering the MoBE.

Acceptance

Deliberately introducing or maintaining specific microbial communities in occupied buildings raises acceptability and liability questions for which no current framework provides adequate guidance, particularly for engineered microbes. Building occupants may vary considerably in their openness to deliberate introduction of microorganisms, which Pillars 2 and 3 encourage. This is especially true in settings with immunocompromised individuals, such as in hospitals, where doctors and nurses may be loath to introduce microbes of any kind given the well-justified historical pursuit of sterility in these environments. These may be the very settings where approaches like PBS could have the most impact, however, given their susceptibility to repeat colonization of pathogens and the importance of HAIs. MoBE engineering approaches can only advance with a good understanding of what is acceptable to building occupants.

Regulation and Alignment with Building Standards

The building industry has made significant strides in developing standards that aim to promote healthy buildings, including addressing microbial exposures. ASHRAE, the WELL Building Standard, and LEED certification frameworks have all begun to incorporate biological quality metrics, but these remain largely limited to mold and endotoxin measurements, and pathogen exclusion via ventilation, rather than more comprehensive approaches to the MoBE. But, these initiatives represent a starting point for thinking about how to extend beyond aerosols and mold to include surface and plumbing-associated microbial hazards, and for adding indicators of community state (diversity, resilience) that are currently absent from all major building performance standards.

Relevant to Pillars 2 and 3, there is no established regulatory pathway for deliberate microbial inoculants in occupied buildings comparable to existing standards for food-contact surfaces or probiotic food products, and the stakes are higher for introducing microbes into occupied spaces compared with, say, bioremediation of contaminated sites. Historically, introductions of presumed innocuous agents have not always been as benign as expected. Regulatory engagement will be critical to mainstreaming MoBE engineering, as safety must remain the central focus of our approach to managing microbes even as we evolve our understanding of what we mean by safe and healthy in this context.

The Healthy Building Movement, calling for enhancing our contact with nature to sustain health, and proposals for “re-wilding” our living environments collectively articulate a vision for the MoBE as a managed ecosystem that has the potential to benefit people. This is a fundamental change to the ways we have approached microbes in buildings, as mainly pathogens to be eliminated. MoBE engineering would seek to maintain pathogen control, but also work to support immune development through appropriate microbial exposures and preserve community resilience; these goals are a logical extension of a broader movement that increasingly views buildings as having the capacity to improve health and well-being.

Moving Forward

We propose a working definition of a healthy microbiome of the BE: pathogens below tolerable risk, non-sterility, and resilient diversity. Each of these pillars is measurable and supported by both ecological theory and converging empirical evidence in the BE, though we note that the pillars differ in the maturity of their evidentiary bases. Pillar 1 rests on well-established QMRA methodology and insights from more than a century of environmental health microbiology. Pillar 2’s health relevance relies primarily on epidemiological associations whose causal mechanisms remain incompletely characterized. Pillar 3 follows from robust ecological theory with emerging empirical support. Taken together, they are principles for defining a healthy MoBE that can be operationalized in engineering these complex systems via feedback-control and integrated into outcome-based building standards. The framework and the relative emphasis of the three pillars should be applied in a context-specific manner.

Our path from this conceptual definition to validated quantitative standards is complex and will require consensus-building in the MoBE community. It will require advancing rapid, quantitative, and viability-indicating methods; more fully developed QMRA frameworks; agreed-upon benchmarks for diversity; new thinking around microbial controls that can preserve beneficial microbes; and epidemiological evidence for validation. The creation of this effectively novel infrastructure for managing the MoBE will require sustained investment by the research, regulatory, and private sector communities of practice who seek to improve the health of buildings and their occupants.

The MoBE field is now at an inflection point where we know enough about these systems to hope to engineer them for our collective benefit, but much remains uncertain. We invite the broader indoor microbiology, environmental engineering, and public health communities to engage with this framework, to critique it, extend it, and test it empirically. Any proposed definition of a healthy MoBE should be judged by its utility.

Biography

Joe Brown, PhD, PE, is the Don and Jennifer Holzworth Professor in the Department of Environmental Sciences and Engineering at the University of North Carolina at Chapel Hill. He is the director of the UNC Water Institute and serves as deputy director of the NSF-funded PreMiEr Engineering Research Center. His interests include novel approaches for engineering the MoBE, methods in environmental health microbiology, quantitative microbial risk assessment, and environmental epidemiology.

The authors declare no competing financial interest.