Bad air can also kill: residential indoor air quality and pollutant exposure risk during the COVID-19 crisis

International Journal of Environmental Research and Public Health Article Bad Air Can Also Kill: Residential Indoor Air Quality and Pollutant Exposure Risk during the COVID-19 Crisis Samuel Domínguez-Amarillo 1, Jesica Fernández-Agüera 1,* , Sonia Cesteros-García2and Roberto Alonso González-Lezcano 2 1Instituto Universitario de Arquitectura y Ciencias de la Construcción, Escuela Técnica Superior de Arquitectura, Universidad de Sevilla, 41014 Sevilla, Spain; [email protected] 2Escuela Politécnica Superior, Universidad San Pablo-CEU, Montepríncipe Campus, Boadilla del Monte, 28668 Madrid, Spain; [email protected] (S.C.-G.); r[email protected] (R.A.G.-L.) *Correspondence: [email protected] Received: 10 July 2020; Accepted: 29 September 2020; Published: 30 September 2020   Abstract: During the first outbreak of the SARS-CoV-2 pandemic the population, focusing primarily on the risk of infection, was generally inattentive to the quality of indoor air. Spain, and the city of Madrid in particular, were among the world’s coronavirus hotspots. The country’s entire population was subject to a 24/7 lockdown for 45 days. This paper describes a comparative longitudinal survey of air quality in four types of housing in the city of Madrid before and during lockdown. The paper analysed indoor temperatures and variations in CO 2 , 2.5 µ m particulate matter (PM 2.5 ) and total volatile organic compound (TVOC) concentrations before and during lockdown. The mean daily outdoor PM 2.5 concentration declined from 11.04 µ g/m 3 before to 7.10 µ g/m 3 during lockdown. Before lockdown the NO 2 concentration values scored as ‘very good’ 46% of the time, compared to 90.9% during that period. Although the city’s outdoor air quality improved, during lockdown the population’s exposure to indoor pollutants was generally more acute and prolonged. Due primarily to concern over domestic energy savings, the lack of suitable ventilation and more intensive use of cleaning products and disinfectants during the covid-19 crisis, indoor pollutant levels were typically higher than compatible with healthy environments. Mean daily PM 2.5 concentration rose by approximately 12% and mean TVOC concentration by 37% to 559%. The paper also puts forward a series of recommendations to improve indoor domestic environments in future pandemics and spells out urgent action to be taken around indoor air quality (IAQ) in the event of total or partial quarantining to protect residents from respiratory ailments and concomitantly enhanced susceptibility to SARS-CoV-2, as identified by international medical research. Keywords: indoor air quality; COVID 19; European; dwellings; buildings; TVOC; PM 2.5 ; CO2concentration 1. Introduction On 11 March 2020, the World Health Organization elevated the categorization of the public health emergency caused by COVID-19 to an international pandemic [ 1 ]. Since the COVID-19 pandemic began [ 2 ], after the first positive case detected in Spain on 30 January 2020, the number of cases detected increased exponentially until the Spanish government approved a Royal Decree 463/2020 on 14 March, declaring a state of emergency for the management of the health crisis situation caused by COVID-19 [ 3 ], forcing the closure of all non-essential economic activities, schools, and sports activities, as well as a ban on travel in Spain. During the state of alarm, people were only allowed to travel on Int. J. Environ. Res. Public Health 2020,17, 7183; doi:10.3390/ijerph17197183 www.mdpi.com/journal/ijerph Int. J. Environ. Res. Public Health 2020,17, 7183 2 of 33 public roads to complete basic activities, with the obligation to stay at home for the rest of the time. Figure 1and Table 1show the chronology of the different phases of the state of alarm in Spain. Figure 1. Chronology and periods of the state of alarm in Spain in 2020. Table 1. Chronology situation of the confinement in homes in Spain in 2020. x Start Date Progression Order Schools closed (SC) 9 March Madrid-region stopped educational activities Pandemic declaration 11 March World Health Organization State of alarm 15 March Lockout Stay-at-home (SAH) order 1st extension period 30 March Total lockdown with stoppage of any activity except essential sectors Stay-at-home order reinforced 2nd extension 10 April As initial stage Stay-at-home order 3rd extension 26 April As initial stage + Children walks allowed (1 h/day) Stay-at-home order 2 May Confinement relief-measures for general population Stay-at-home order 4 May Resumption (limited) of commercial activities Onset of deconfinement 4th extension 10 May Deconfinement phases 5th extension 24 May Deconfinement phases 6th extension 7 June Deconfinement phases 21 June End of state of alarm Given this situation, millions of Spanish citizens remained confined to their homes until the beginning of the de-escalation phases. This confinement meant that the population, which previously did not spend more than 60% of their daily time in the home, were now spending 100% of their time in the home on a continual basis. The city of Madrid was one of the countries hardest hit by the coronavirus, both in terms of the number of people infected per million inhabitants and the total number of deaths. The mortality rate in Spain was 54.73 deaths per 100,000 inhabitants; in Madrid it was 312 deaths per 100,000 inhabitants on 5 May 2020 [4,5]. Figure 2shows the mortality from all expected and observed causes in the community of Madrid (blue shading represents the average evolution of the previous 10 years). Various studies on the emotional impact of the coronavirus episode in the first days of confinement in Spain [ 6 , 7 ] explored the relationship between a comprehensive set of lifestyles, sociodemographic variables, and personality and mood variables during the early stages of quarantine. The presence of low tolerance to negative emotions, neuroticism, and, to a lesser extent, friendliness, sleep quality, the presence of children, and time spent surfing the Internet presented the strongest correlations with the population’s moods. A similar study was conducted in Portugal, in which the lifestyle habits, anxiety levels, and basic psychological needs (BPN) in Portuguese adults during the COVID-19 pandemic were characterized, including a comparison between sex and age groups [ 8 ]. Other studied parameters of interest include the population’s activity level and anxiety [9], and the relationship with addiction [10]. Int. J. Environ. Res. Public Health 2020,17, 7183 3 of 33 Figure 2. Mortality from all expected and observed causes in the community of Madrid (blue shading represents the average evolution of the previous 10 years). The measures implemented by the governments of different countries have caused numerous environmental impacts, some of which can be considered positive, such as the improvement in air and water quality in urban areas, whereas others have been negative, such as the deterioration of the quality of coastal effluents due to the elimination of sanitary supplies [ 11 ]. The improvement in air quality has been particularly notable in large cities in China, where outdoor pollution levels (NO 2 and PM 2.5 ) have decreased significantly during lockdown [ 12 , 13 ]. In European cities like Barcelona, NO 2 concentrations were reduced by half during the lockdown, PM decreased lower, and O 3 concentrations increased by around 50% [ 14 ]. PM 2.5 deterioration of air quality may lead to various kinds of pulmonary and cardiovascular diseases causing premature death [ 15 ]. In addition, there are several studies that indicate cities with poorer outdoor air quality have had a higher death rate in those affected by COVID-19 [16–20]. The reduction in environmental pollution in cities is a positive aspect for their inhabitants, which can contribute to improving living conditions during confinement by allowing for healthier ventilation. During this period, most of the population has been exclusively confined to their homes, so they were exposed to the same indoor atmosphere almost 100% of the time. In many cases, these homes are not entirely suitable for intensive use, with little interior space in relation to the number of inhabitants and, on many occasions, without adequate systems for thermal comfort, ventilation, and indoor air renewal [21–23]. Although the most common type of housing in Europe is single-family, this situation is different in Southern Europe, where collective residential apartment buildings predominate [ 24 ], especially in social housing [ 25 ] and in the larger proportion of social housing in the current general housing stock [ 26 , 27 ]. Although with regional variations, the architecture and construction of multi-family buildings are similar throughout the area [ 28 ], differing from the usual solutions in central and northern European countries. Many of these architectural solutions are a legacy of the residential construction boom of the 1960s and 1970s [ 26 , 28 ]. These southern European inhabitants also share cultural and social similarities both in the way people live in their homes and in managing their ventilation, heating and cooling system [21,29,30]. Household users are exposed to several pollutants. There are numerous source of pollution produced by the occupants and their activities, by the furniture and building materials, by technical defects, the presence of humidity or dirt, and, in particular, the use of chemicals (cleaning, hygiene, and personal care products) as well as tobacco smoke in smokers’ homes [ 31 – 34 ]. Pollution from outside such as pollen, dust, and industrial and traffic pollution must also be considered. Through a review of the literature, Lozano and Siegel provided evidence for most of the countries studied of the Int. J. Environ. Res. Public Health 2020,17, 7183 4 of 33 prevalence of poor environmental quality conditions along with a lack of thermal comfort in social housing units [35]. There are numerous studies on indoor quality and sick building syndromes, especially in cold countries [ 36 – 41 ], but studies in temperate areas of Southern Europe are limited, with some studies in operational housing conditions [ 42 – 47 ]. Studies of the indoor air quality in homes during the COVID 19 lockdown are also limited, although some studies point to household cleaning products as a particularly relevant source of indoor pollution. Many people may be cleaned more frequently and with stronger disinfectants to reduce the potential of viral infection. Efforts to make homes airtight to improve energy efficiency have created buildings with reduced outdoor ventilation rates resulting in the buildup of indoor pollutants to harmful levels that would be otherwise unacceptable outdoors [ 48 ]. The aim of this study was to determine the impact of COVID-19-related confinement on the indoor air quality of dwellings in Madrid. The concentrations of CO 2 , PM 2.5, and TVOCs were measured over time before and during lockdown in relation to the thermal-physical parameters of indoor air quality to identify health risks of the occupants due to this continuous exposure and the alteration of the usual cycles in the dwellings. 2. Materials and Methods 2.1. Sampling Madrid is the capital of Spain. The population of the city is about 3,182,981, although if the floating population and the metropolitan area are considered, this number exceeds 6,685,471 inhabitants [ 49 ]. Located at 40 ◦ 41 0 and an altitude of 659 m above sea level, its climate is characterized by a transition between the cold semi-arid (Köppen BSk) and the Mediterranean (Csa), with an average annual temperature of 15 ◦ C. Winters are moderately cold, with average temperatures in the coldest month (January) of around 6 ◦ C. Summers are hot, with the average temperature exceeding 25 ◦ C in July, with average maximum temperatures of between 32 and 33.5 ◦C. The daily temperature oscillation is significant in the city periphery (around ∆ t 13 K), but is less so in the city urban area (daily ∆ t goes usually below 10 K). The annual thermal amplitude is also high: between 19 and 20 ◦ C. This climatic context traditionally favored the use as naturally-ventilated houses promoting a semi-open state with most windows open for a significant portion of the year, a trend that is declining in the southern European cities due to outdoor pollution and other factors, such as noise and behavioral pattern changes. This situation makes the area especially relevant as a typical situation for analyzing the behavior of the household stock, which shared many building and behavioral characteristics with other southern European urban agglomerations during the pandemic. The most widespread residential buildings are multi-family dwellings, either in linear or H-shaped blocks [50]. For this work, four case studies representative of the typical legacy social housing in the city, were selected for monitoring their environmental conditions under actual operational conditions before and during the state of confinement. Figure 3shows the case study typologies. The typical construction of these buildings—mainly brick masonry and reinforced concrete—results in very compartmentalized housing units, with the apartments as unitary volume units. Generally, the buildings lack common elements that allow air communication between units (there are typically no chimneys, or common ventilation ducts). Previously conducted assessments [ 51 ] showed that the air leakage ratios to adjacent units did not exceed 4% of the air flow (tests at 50 Pa pressurization). For practical purposes, it is possible to consider the apartment units as independent from each other. The selected case studies represent the typical legacy social housing in the city plus one contemporary development. Different profiles of occupants with enough representativeness of the structure of the use of national dwellings were selected. The case studies were selected due to the characteristics of the building and the family composition, to provide a sample of the most common types of families: from a single adult living alone (CS3), which represents 25.3% of the population of Int. J. Environ. Res. Public Health 2020,17, 7183 5 of 33 Spain between the elderly and under 65s living alone, a couple with one child (21%), a couple with two children (17.6%), and a couple with two children and a dependent elderly person (5.7% of Spain’s total but growing quickly). Figure 3. Case study typologies. The example selection was developed from the INFILES National R&D Project (on residential buildings infiltration national assessment) [ 52 , 53 ], due to the representativeness of national housing (the prototypical characteristics, both in morphology and construction and equipment). The daily hours of occupation refer to the average occupation of each one of the groups of inhabitants before the stay-at-home order—business as usual occupation—and when the lockdown period was in force (figures in brackets) when the inhabitants mostly remained at home (apart from essential workers, the population could only leave home to make quick purchases of basic items) almost continuously until the beginning of the deconfinement. 2.2. Monitoring The monitoringperiodran from1February to30April 2020. Indoordata(thermal and pollutants)were recorded at an interval of 15 min using electronic sensors and a data logger system. Detailed information on the characteristics of the measurement equipment is in Appendix A. Thermal ambient monitoring was conducted in accordance with ISO7726:1998 specifications (ISO/TC 159/SC 5, 1998), both in terms of the instrumentation and methods [ 42 ]. The system was in continuous operation during both time periods. The outdoor environmental data were provided by the Air Quality Service of Madrid Council (Station 28,079,007 of the Air Quality network). The outdoor air quality station was particularly representative both for its location, barycentric to the cases in the urban center and for the reliability of the measurements. The case studies were located around the station with distances ranging from 2 to Int. J. Environ. Res. Public Health 2020,17, 7183 6 of 33 4.33 km from it, all of which are located within the consolidated urban area of the city (not in the suburbs). The variations with respect to other stations in the area were not very significant in general. The general climate data were provided by the Spanish national meteorological office (AEMET) [ 54 ]. Portable reference sensors were used to record periodic contrast measurements in situ to ensure the consistency and accuracy of the measurements of the monitoring systems. 2.3. User Information Gathering A survey was conducted on the housing equipment management and habits, which was complemented by a telematic interview with home inhabitants. The survey asked about their home profiles (number and schedules of inhabitants), the heating, air conditioning and hot water systems provision and use, main appliances, as well as the use schedules before and during lockdown (as shown in Table 2). Table 2. Characteristics of the dwellings in the study sample (numbers in brackets represents any changes during lockdown). Case Study CS1 CS2 CS3 CS4 Year of construction 1954 1950 1950 2008 Floor area (m2)78.5 74 73 90 Free Height (m) 2.93 2.74 2.8 2.7 No. bedrooms 3 3 1 3 No. bathrooms 2 2 1 2 No. Occupants 3 4 1 5 User profile Couple with children Couple with children 1 adult Couple with children and grandfather Retrofitting Yes Yes Yes No Hours per day in the house-occupant 1 Weekday 12 (12) 20 (24) 12 (24) 12 (24) Weekend 16 (24) 15 (24) 15 (24) 19 (24) Hours per day in the house-occupant 2 Weekday 16 (24) 20 (24) 12 (24) Weekend 22 (24) 15 (24) 19 (24) Hours per day in the house-occupant 3 Weekday 12 (24) 20 (24) 10 (24) Weekend 22 (24) 15 (24) 19 (24) Hours per day in the house-occupant 4 Weekday 14 (24) 10 (24) Weekend 15 (24) 19 (24) Hours per day in the house-occupant 5 Weekday 19 (24) Weekend 18 (24) Heating system Central-heating Individual boiler Central-heating Individual boiler Water radiator Water radiator Water radiator Water radiator Cooling system Split XD unit in all house Split XD unit (bedroom and living room) no portable evaporative-cooler n50 (h−1): airflow leakage at 50 Pa 4.87 1.2 3.3 8.22 Smokers yes no no no Gas stove no no no no Domestic Hot Water Boiler (indoor) electric gas electric gas Urban location Downtown residential street Residential street with green areas Close to a main traffic route and green areas Residential street The interviews were focused to study both the way the apartment was used and the user’s personal perceptions of the habits. The inhabitants were asked regarding their habits before and during confinement including alterations or perceived changes: a) Household cleaning: cleaning habits, use of cleaners, disinfectants, product types changes, as well as the use of air fresheners. b) Voluntary ventilation (manual window operation): how they performed ventilation, window opening times and frequency. Int. J. Environ. Res. Public Health 2020,17, 7183 7 of 33 c) Laundry: estimated washing frequency, hanging and ironing location and type. d) Confinement management: home entrances and exits, the frequency and number of people and special behaviors during confinement. e) COVID hygienic special actions: entry/exit cycles procedure, disinfection processes, treatment of purchased products, etc. f) Tobacco: smokers in the house, daily consumption, smoking place, and related habits. 3. Results 3.1. Exterior Ambient Quality As the general purpose of this study was the longitudinal analysis before/after the homes under study due to the need to keep its occupants in them 24/7 and the effects that this process produced on the interior atmosphere. The analysis of the current situation was conducted to gather a general knowledge of the boundary conditions of the homes beforehand and during confinement. While a specific analysis of the evolution of the city atmospheric quality in this period requires a specific and detailed analysis work, which is out of the scope of this work, it is of interest to have a general overview of the outdoor situation in both the prior period and during the main period of confinement. In general, the emergence of COVID-19 significantly influenced the emission of atmospheric pollutants, especially those linked to road traffic. According to information from the Madrid City Council, the closure of schools caused a drop in traffic of around 8.5%. The subsequent confinement, in which work activity was allowed, produced a significant decrease in traffic intensity, which was between 66% and 70% of a working day equivalent. This was somewhat more noticeable on the weekends, where it reached a reduction of up to 78% reduction. The period of limiting activity to the strictly essential, however, had less of an impact, with an additional drop of only 6% in vehicle movement over the previous week [55]. The external environmental situation of the urban environment significantly improved in all indices, mainly due to the drastic reduction in road traffic and public transport. However, the heating systems of the city buildings (mainly RSI: residential and services buildings) remained a significant source of pollution at the urban level—the average second source of pollution of the city with around one-third of the NOx and PM contributions [56]. Those systems maintained a base level of the most common external pollutants, in particular in the most intense phase of confinement at the end of March and beginning of April when the outdoor temperatures were the lowest of the period, which was fundamentally reflected in the NOx and PM as can be seen in Figures 4and 5, although with an oscillating profile closely linked to the weather conditions. Figure 4shows the concentrations of PM 2,5 and PM 10 , outside in Madrid before and during home confinement. Figure 5shows the concentrations of NO, NO 2 , and NOx, outside in Madrid before and during home confinement. For a general characterization of the city’s atmosphere, we can select, as the main indicators, the PM and NO 2 values as key parameters of the Air Quality Index (AQI) National System [ 57 ], which provides continuous information to the citizens on environmental quality. In general, the situation of particulate matter in the period analyzed was acceptable to good for the usual standards. There was an average PM 2.5 daily mean of 11.04 µ g/m 3 in the pre-confinement situation—an AQ index of good (11–20 µ g/m 3 )—with only some days rising up to 24 µ g—a regular AQI (21–25 µ g/m 3 ). This situation evolved to a significant reduction during the main period of confinement, with the average of the daily means set to 7.10 µ g/m 3 —an AQI of very good (0–10 µ g/m 3 )—with only a few days (8.88%) rising above 10 g/m 3 (maximum of 13 µ g/m 3 ), which is a very suitable condition with a reduced presence of fine particulate matter in the city environment. With regard to the value of PM 10 , the influence on the period was even smaller with most days located in very good ranges—an AQI of very good (0–20 µg/m3) and only a few days prior to confinement with values up to 32 µg/m3. Int. J. Environ. Res. Public Health 2020,17, 7183 8 of 33 Figure 4. Concentrations of PM 2.5 and PM10 outside in Madrid before and during home confinement. (Source Sistema Integral de Calidad del Aire de Madrid). Figure 5. Concentrations of NO, NO 2 , and NOX outside in Madrid before and during confinement. (Source Sistema Integral de Calidad del Aire de Madrid). The overall situation regarding NO 2 was similar, with an initial stage at acceptable values for the typical city period, with average hourly values usually in the very good band (46.73% of the period) or good (51.79%)—the AQ Index ranged from 0–40 and 41 to 100 µ g/m 3 with 1.49% of the hours above 100 µ g/m 3 but not surpassing the 110 µ g/m 3 value, and far from the limit of 200 µ g/m 3 set by European legislation [6] as the health limit value. Int. J. Environ. Res. Public Health 2020,17, 7183 9 of 33 During confinement, the situation evolved to very good in a generalized way (AQI 0-40 µ g/m 3 ) with the 90.93% of the hours within this range and only 9.07% in good condition. The peak value was 64 µ g/m 3 . With regard to NOx, while this were not included in the city’s air quality indicator system and there is no regulatory limitation for their values, it is equally of interest to comment on their values during both periods. Prior to confinement, as in the NO 2 case, their values were typically relatively contained with an average hourly average of 66.46 µ g/m 3 without reaching the alert threshold situation (average hourly concentrations of more than 400 µ g/m 3 maintained for 3 h), with only two point short peaks reaching 320 µg/m3. In the confinement phase, these values became quite small with an average hourly average of 25.77 µ g/m 3 and a maximum peak of 89 µ g/m 3, far from the alert thresholds. Although the indoor direct effects of these species are not included in this work, an analysis was performed to verify that no special or singular situations occurred in the outdoor atmosphere that could have generated anomalies in the previous periods or during confinement. The average values tended to decrease as the month of April progressed, and the outdoor temperatures rose, which, with the traffic contained by the restrictions, is linked with a reduction of the heating uses. As complementary information, the tropospheric urban ozone levels were low to very low, for both periods, mostly due to the climatological conditions. An average 8-h mean of 41 µ g/m 3 prior to the confinement and 57 µ g/m 3 in the lockdown time were measured, figures that are below the target value for the protection of human health (120 µ g/m 3 , eight-hour average) as in DIRECTIVE 2008/50/EC [ 58 ]. The atmosphere situation with regard to the ozone concentration allowed for the very good or good classification (0–80/81–121 µ g/m 3 hourly mean) using the Air Quality Index of the city during both periods, prior and during confinement, with all the hourly means under the 100 µ g/m 3 figure and below the legal public information threshold (180 µ g/m 3 hourly). Therefore, for practical purposes and in terms of longitudinal analysis (before/after) given these values, we considered that the presence of tropospheric ozone was not very significant in the modification of the interior conditions of homes due to confinement. In the case of having occurred during the summer, we would expect these figures to be higher, given the city’s historical records. 3.2. Information from Users Case study 1 has the only person who was active outside the home during confinement—being considered an essential worker. In this case, while there was an increase in the working hours, they had an irregular profile. Working-age adults in case studies 2, 3, and 4 maintained their activity during confinement through telework, leading to an increase in the use time for computers and printers. The routines of entering and leaving the home were severely restricted. In general, only one person from each household was the one who habitually left. Except for case 1 in the work periods, the rest were specific outings of very short duration for the acquisition of basic supplies or pet walks (as established by the regulation of the alarm state). The rest of the inhabitants remained in the house without leaving it during the entire period. The routine for entering and leaving the home is shown in Table 3. Table 3. Confinement home entry/exit routines. Activity CS1 CS2 CS3 CS4 Work out-of-the-house daily (1 person) no no no Telework no 2 people 1 person 2 people Supplies daily (1 person) each 3 or 4 days daily each 3 or 4 days Pet walking 3 times/day - - - During confinement, all cases reported an increase in the time of use of television and other electronic entertainment equipment (video consoles, tablets, audio systems, etc.). One of the most significant aspects reported was the intensification of household cleaning, also affecting households without regular exits/entrances. Int. J. Environ. Res. Public Health 2020,17, 7183 16 of 33 3.3.3. PM2.5 Airborne particles are solid or liquid substances suspended in the air stream, and can include dust, fumes, smoke, microorganisms, mists, and fogs [ 70 ]. These particles can vary widely in diameter, with the most harmful in the domestic environment being respirable suspended particles (RSPs), especially the fine inhalable particle fraction sized 2.5 µm or less (PM2.5) [71–74]. The World Health Organization (WHO) has established reference target values both for long-term exposure concentration with annual means under 10 µ g/m 3 and short-term guidance for the 24 h means recommended to be under 25 µ g/m 3 at the 99th percentile (3 days/year) [ 75 ]. The European limit value for the annual average is higher at a 25 µ g/m 3 limit. The outdoor concentration of PM 2.5 for Madrid did not present significant variations between the initial period and the lockdown except for specific peaks, with a daily average around 6 µ g/m 3 , although some decrease was observed during the reinforced lockdown (1st extension period). Therefore, the external situation can be understood as relatively steady. These figures are also typical for the city as it is one of the few European cities that consistently stays below the EU and WHO recommendations [ 76 ]. However, there is a general consensus on the need to keep the levels of exposure to PM 2.5 as low as possible, as it has not been possible to establish a minimum safety threshold without health effects, especially inside buildings [ 77 – 79 ]. The reduction of exposure has especially positive effects on sensitive inhabitants, such as individuals with respiratory conditions, the elderly or children [ 80 , 81 ]. Likewise, there appears to be evidence that exposure to PM of internal origin is more aggressive in the respiratory effects than of external origin, with a greater capacity for lung response [ 82 , 83 ]. As a reference, the typical mean residential indoor concentrations in Europe range from 7 (Helsinki) to 21 µ g/m 3 (Athens) [ 84 ] and up to as high as those in Italy with 50 µ g/m 3 by Gotschi et al. [85]. Although there are innumerable outdoor sources, there are also internal sources responsible for the emission of particles, which can assume a special role in situations of high intensity use of indoor spaces, such as during a confinement situation. One of the main sources of particles inside residential buildings is cooking, as food preparation causes the emission of vapor, smokes, and aerosols [ 86 , 87 ] and smoke and effluents can be released indoor if there is a lack of a suitable extraction system and by indirect paths due to the possibility of gas-reentry through the infiltration airflow [ 88 , 89 ]. While not as common in residential buildings as they are in offices and public buildings, the presence of printers may also contribute to particle emissions—they are a significant VOC source [ 90 ]; however, due to the wide spread of teleworking and tele-teaching during confinement, their presence increased significantly in homes, as well as their use, so they must be taken into account in the analyses. Another large contributor, which is increasing in homes today, is the use of cleaning agents [ 91 , 92 ], as well as cosmetic and personal hygiene items (deodorants, hair sprays, etc.) [ 93 ]. Although these are not typically taken into account, the use of electric air fresheners also has a high capacity for emitting particles [ 94 – 96 ]. Since airborne particles are deposited on adjacent surfaces, one of the aspects with the greatest influence on the increase in indoor particle concentrations is the re-suspension due to the movement of the occupants. [ 88 ]. This happens very frequent and intensely in indoor environments with regular movement and a low ratio between volume and inhabitants, such as is the case in homes [ 97 , 98 ]. In addition to the emissions caused by the cleaning products, the cleaning process itself, with surface cleaning, sweeping, and other motions, has been identified as an activity responsible for the resuspension of the particles and, therefore, of increased concentrations in the air [99]. In general, the indoor averages of both periods were higher than the city reference outdoor values, as shown in (Figure 11). In the period prior to confinement, the dwellings showed an average between the double (CS4) and triple (CS3) of those found outdoors. During confinement, the averages increased, except in CS4, with central values more than three times the external ones. This issue highlights the significant contribution of indoor sources of emissions, a situation that intensified during confinement. Int. J. Environ. Res. Public Health 2020,17, 7183 17 of 33 Figure 11. The average PM 2.5 concentration and airflow leakage at 50 Pa (n 50 ) for the case studies for the period before and during the lockdown in Madrid. The higher or lower air-permeability of the envelope did not appear to play a significant role in the indoor PM 2.5 concentrations through the fabric particle penetration process, as there were high levels of concentration both in airtight homes (CS2) as in average permeability homes (CS 1). Nevertheless, the less airtight dwelling (CS4)—beyond the stock average value [ 61 ]—was the one that typically had the lowest PM 2.5 concentrations. In this case, the airtight homes could benefit from greater air-dilution through the base infiltration, which would contribute to maintaining greater control over the indoor concentration when there are scenarios of limited indoor emissions unlike in CS1. These questions, although require a more comprehensive and in-depth study, which shows a line of interest for future work, and does seem to indicate the prevalence of internal source dynamics over external source dynamics in this situation. Adequate information on the evolution of particulate matter can be obtained through daily averages, representative of short-term exposure. The representative values before and during confinement are shown in Tables 9and 10, represented by the medians and their associated statistics. The typical 24-h average values were initially above 10, although some days had values rise to 57.72 µ g/m 3 . In general, as was more evident in cases 1 and 2, the inter-days variation was high, alternating days of low emissions with higher ones, a situation linked to the routines of these households—regular and irregular cycles. Table 9. The daily PM2.5 means (µg/m3) before lockdown. Case Median SD Variation Coefficient Maximum CS1 14.60 8.91 57.89% 57.72 CS2 12.84 12.21 72.90% 56.05 CS3 16.07 9.14 59.27% 41.43 CS4 10.63 4.77 42.66% 27.64 Table 10. The daily PM2.5 means (µg/m3) during lockdown. Case Median SD Variation Coefficient Maximum CS1 16.37 9.55 49.97% 60.89 CS2 15.47 14.16 74.10% 90.43 CS3 16.94 8.51 53.14 47.27 CS4 7.19 3.22 40.67% 16.73 Int. J. Environ. Res. Public Health 2020,17, 7183 18 of 33 During lockdown, the previously identified values rise was identified in detail as a general increase in the daily average values except for CS 4, with an increase of 12% to 20% in the daily intensity. Likewise, the intra-day variability increased, with low and high emission cycles, reaching mean values of 24 h as high as 90.43 for CS 2. This increase can be associated with two basic aspects: the increase in cooking with regard to the previous conventional routine, which set the base values as is repeated in 24-h cycles, as well as by the intensification of domestic cleaning and housekeeping tasks with respect to the normal situation, which was not developed daily but on separate days. Specific short-time high-peak emissions events were identified in all the homes, although they were not representative (values went up to 1000 µg/m3in certain cases). Although the period-means (prior/lockdown time) are referred to as a median-term exposure (around 45 days for each set of measures) those figures are above the annual recommended target for WHO. While a direct comparison is not possible, this may indicate a potential risk of exposure if forecasted over time. Of greater interest is the comparison of the WHO limits for the 24-h averages (25 µg/m3), which were exceeded for a significant number of days in each period. In the normal situation, the percentage of days when this threshold was surpassed were: 4.65%, 16.6%, 53.3%, and 2.32% (for cases 1 to 4) and during the confinement these figures increased to 16.6%; 22.9%, and 5% for cases 1 to 3, and CS 4 did not exceed the values on any day in the period. As the epidemiological evidence shows the adverse effects of PM are linked to both short-term and long-term exposures [ 100 ]. In this case, the apartments started from an initial profile that was not good—in this period, the actual exposure was lower as the occupation time was also shorter—that worsened during confinement, increasing the risk for the occupants of the dwellings, against the short-term exposure levels and related to the inhabitants—for instance, CS1 and CS2. As indicated above, in the analysis of the temporal 24-h profile of particles emissions (PM 2.5 ) inside the homes, the peaks were associated with the periods of food preparation and daily cleaning, and, as shown in Figure 12, the measures are represented in relation to their temporary presence during the confinement period. Notably, in Southern Europe, the time for dinner is later than in Central Europe, an aspect that seems to have intensified during the confinement period. Figure 12. Hourly evolution of the PM2.5 of the case studies during the lockdown in Madrid. 3.3.4. TVOC Along with particles, the most common species in the home environment are volatile carbon compounds (VOCs), represented in this case by TVOCs. Nearly 900 VOCs have been identified in the indoor environment [ 101 ], mainly related to interior furniture, finishes, and other construction elements, such as paints, solvents, adhesives, carpets, fabrics, and textiles, but above all to the presence of occupants in the dwelling and their activities, as well as the effect of increased turbulence in the Int. J. Environ. Res. Public Health 2020,17, 7183 19 of 33 indoor atmosphere caused by movement within the different enclosures (increasing the effect of mixing the pollutants). In urban areas, there are also external sources of VOCs, particularly those related to traffic as well as other combustion equipment. There is evidence that, in European residential and service buildings, the VOC levels are typically higher indoors rather than outdoors—often several times higher [ 102 – 104 ]. Likewise, the action of the outside environment, either through ventilation or air permeability, will manifest in a greater or lesser dilution of this component indoors [105]. Exposure to high concentrations of VOCs in air can result in a variety of toxicological effects, including fatigue, headaches, drowsiness, eye irritation, and many others [ 106 , 107 ]. However, the wide variety of potential sources and compositions makes it impractical to measure concentrations of each chemical individually. As such, the concept of total volatile organic compound (TVOC) attempts to address this practical limitation by providing a simple measure of the aggregation of all VOCs without distinguishing between individual chemicals. Although no unified regulation exists for exposure to total VOCs within the home, some general thresholds can be assumed as a result of physiological responses of and disturbances to the occupants. In the range of 120 to 1200 ppb, symptoms of irritation and discomfort may begin to appear in some occupants. Above 12,000 ppb, this discomfort is generalized, with different manifestations depending on the degree of sensitivity of the inhabitants and the species present, with toxicity developing from 10,000 ppb [ 108 ]. As reference levels, <120 ppb is considered a healthy environment or VOC-free and >1200 ppb is considered a risk environment with intermediate figures assumed as acceptable but not a risk-free situation. These considerations are general in nature and are highly variable depending on the type of VOC present in the atmosphere [ 109 ]. The typical average outdoor levels of TVOC for Madrid (urban center) fluctuated around 14 to 2.85 ppb, although there was a large time variability and weather condition dependency [ 56 ]. The most representative species in the city atmosphere were toluene and benzene [ 14 ]. During the period prior to confinement, the average daily concentration was Tol: 0.86 and Ben: 0.343 ppb (max. extreme day of 1.75/0.50 ppb). These concentrations shifted to a significant reduction during the period of confinement, going down to a daily average of Tol: 0.146 and Ben: 0.085 ppb (extreme day of 0.39/0.188 ppb) for the last 15 days from March to the end of April. Therefore, there was a significant drop in the VOC outdoor air presence, both on average and extreme days, which may be associated mainly with the drastic reduction of city traffic. This highlights that, during the confinement period, the influence of the outer species, although present in the interior, were not the main drivers of the high interior levels and the evolution observed between periods. In the case of TVOCs, dwelling 1 was different from the rest as it had one smoker and, even if smoking was not conducted directly in the home, this did present indirect effects that affected the interior quality: the smoke residues on clothing, odors, etc. The high values usually found in this dwelling necessitated conducting the study independently from the rest of the sample. Since relatively high extreme values were found in all dwellings, although only occasionally, the value of the median was more appropriate than the average for representing the behavior of the indoor environment, as the latter is more affected by the presence of extreme values. Excluding CS 1, the medians, both prior to and during confinement, were within the typical findings of similar contemporary dwellings [ 42 ], although these were higher than desirable and far from healthy environments. The representative values were within the acceptable label atmospheres, but their high variability, with a variation coefficient of 38.5% to 93.89% for the pre-COVID period and 58.95% to 109.88% for the confinement, indicates periods with high or very high levels of contaminant concentration that became more frequent during confinement. Tables 11 and 12 show, respectively, the statistical data on the concentration of TVOCs (ppb) before and during lockdown. Int. J. Environ. Res. Public Health 2020,17, 7183 20 of 33 Table 11. Statistical data on the concentration of TVOCs (ppb) before lockdown. Case Average Median SD Variation Coefficient Maximum 5% 95% CS1 (*) 5550.11 1983 7982.31 143.82% 60.000 354 21,696 CS2 438.453 429.0 169.153 38.57% 2187.8 179 649 CS3 610.77 510.2 397.087 65.014% 5698.2 146 1.367 CS4 272.524 213.9 255.89 93.89% 3954.1 61.5 661 (*) presence of habitual smokers in the home. Table 12. Statistical data on the concentration of TVOCs (ppb) during lockdown. Case Average Median SD Variation Coefficient Maximum 5% 95% CS1 (*) 15,285 11,090 12,939.9 84.65% 60,000 2.902 46,383 CS2 748.88 591.95 441.54 58.96% 7665.4 338.5 1583 CS3 471.61 401.4 518.22 109.88% 10,530.5 195 876 CS4 438.781 346.2 347.83 79.27% 3964.1 117 1190 (*) presence of habitual smokers in the home. The frequency over time of the different concentration levels for each time period offers an insight to the kind of exposure that each housing is presenting, which can be expressed through the concentration level temporal percentile. In the normal use of homes (prior to confinement) the exposure of users to high concentrations of VOCs was not very significant—except for case 1—and indoor environments usually maintained an acceptable quality threshold of 1200 ppb (92.5 to 99.5 percentile) (Figure 13), with dwelling 3 being somewhat more exposed. This home was also the one with the highest peak values and width of the distribution—as well as the highest STD value (σ3: 397 ppb). Figure 13. Box and whisker plot of the TVOC concentration for the case studies before and during the COVID-19 lockdown in Madrid. This can be associated to the profile with a single-inhabitant home and, therefore, the periods of activity/non-activity showed more drastic differences and had a less smoothed profile than others. However, only Case 4 presented significant good quality periods—considered VOC-free when under 120 ppb, with a percentile of 18% for this threshold. In the other situations (1 to 3), periods of low concentration were unusual—with percentile values between 3.0 and 4.3. The periods of occupation were not continuous, and the dwelling had several unoccupied hours. However, during confinement, the usual concentration increased for dwellings 1, 2, and 4 (increases in the median from 37% to 559%). Excepting Case 1, cases 2 and 4 maintained their central distribution within the acceptable category despite the widening of their standard deviation ( σ2 : 441 and σ4 : 347), although this threshold did not assure risk-free and the exposure increased as the time spent at home did. The time out of the over 1200 ppb threshold reduced in cases 2 and 4 (88% and 95%, respectively). In Case 3, the opposite situation occurred, with a reduction in both the mean and the median (–22%), although with a widening of the standard deviation and a particularly high extreme value peak event. This highlights a changes in the indoor routines—less laundry, less use of cosmetic agents, Int. J. Environ. Res. Public Health 2020,17, 7183 21 of 33 and similar conditions, as described above—and the time below this threshold increased to 97% of the time; therefore, extreme values occurred at very specific moments and for short periods primarily linked to cleaning actions with higher values than before due to more intense actions but with less frequency. Conversely, in all cases, the almost constant occupation of the dwelling caused the practical non-existence of low periods of VOCs (TVOC <120 ppb), a situation that can be observed in its probability curves (Figure 14) and the general increase in the values for the 5% percentile. Figure 14. Quantile graph of indoor TVOC concentration before and during the lockdown. Dwelling 1, with smoking occupants, presented a unique case. VOC levels were usually high, both prior to and during confinement. Its initial median concentration was 1983 ppb with an average of 5500 ppb due to the presence of extreme values, and with a standard deviation much higher than that found in other dwellings ( σ1 : 7982). This indicated both a large width of the central part of the distribution as well as the presence of occasional situations of extraordinarily high values. This dwelling, under a normal situation, usually had an atmosphere with a high presence of VOC’s, being located most of the time within 354 to 21,696 ppb (5% and 95% percentiles), with very limited periods of good quality atmospheres (<120 ppb, 1.7%; <1200 ppb, 39%). However, the temporary presence of the occupants, with daily work cycles, to some extent reduced the exposure to these species. During confinement, the concentration of TVOCs in CS1 increased in the medians, extreme values (maximums), and 95% concentrations. The median increased by an order of magnitude of more than five, raising the average above 15,000 ppb. The peak levels surpassed 60,000 ppb. This indicated a very continuous presence of substances at dangerous levels for both short- and long-term exposures. The imposition of confinement worsened the situation, as the hours indoors increased significantly, with the occupants being at risk of almost continuous exposure. In contrast to the previous situation, hours below 120 ppb were negligible (0.025%) and only 0.45% of the period was expected to have TVOC levels below 1200 ppb. Although smoking related effects were are usually the main factor responsible for indoor peak emissions [ 110 , 111 ] concerning body and clothes residues, as reflected by the periods of activity in the dwelling and the time profile, this caused secondary emissions. To neutralize tobacco odors, in combination with the presence of domestic pets, electric air fresheners were used continuously, generating a continuous emission of VOCs, responsible for the identified high background level as informed by the inhabitants. This situation is compatible with the findings in [ 21 , 29 , 112 , 113 ], where the daily clothes laundry of the worker also contributed to the emissions. The analysis showed a statistically significant difference in the sample distributions between the prior and lockdown situations (at a 95% confidence level) for the contrast tests (T, Mood, Int. J. Environ. Res. Public Health 2020,17, 7183 22 of 33 and Kolmog ó rov–Smirnov test K-S test) for each case (p>0.05 in all cases). The data distribution of each case period, especially those for the confined state, better fit the log-log-type distributions compared with the normal. 4. Discussion The sudden change from a presence in the home of between 12 and 15 h per day to practically 24 h/seven days a week (as inhabitants were not allowed to leave the home except for the purchase of essential supplies and work activities for adults), caused an over-strain on the current dwellings, many of them without the adequate capacity to manage this rise in use intensity or lacking the necessary equipment [21,29]. The at-home confinement situation produced evident alterations in the environmental quality of the dwellings. Starting from typical stock conditions for this type of home [ 42 , 114 ], the confinement increased the exposure conditions, in most of the cases, and, when abnormal starting conditions were present (as in CS1), they worsened. Although this situation was expected, the results should arouse concern in most cases, with the identification of multi-factorial risk factors: VOC and PM concentrations, ventilation strategies and effectiveness, continuous inhabitant presence, and a lack of inhabitant awareness of IAQ as well as other aspects not yet identified but linked with ineffective ventilation. This can affect inhabitants both with possible short- and long-term effects. The multifactorial nature of the exposure and effects presents a scenario that should be considered, both for the analysis of the impact on inhabitants and in anticipation of future confinement episodes for the next wave(s) of this/these infection(s). Notably, together with a significant increase in the sources of internal pollutants, the indoor environmental quality conditions (IEQ) were far from desirable figures. Although the dwellings generally had an acceptable air renewal under the usual standards (combination of infiltration and voluntary ventilation), normal CO 2 concentrations were below the WHO recommendation threshold, with some exceptions. This does not seem sufficient to ensure the dilution of part of the indoor pollutants (mainly TVOCs and PM) to healthy levels, especially when these are emitted in massive and high frequency amounts. The concentration of TVOCs in homes over time can be used as an indicator of the level of risk of their environmental exposure. Over time, the alteration produced by the different stages of confinement can be clearly identified. In all cases, we observed an evident change in the profile, although with a certain time lag depending on the specific characteristics of each family unit, reflecting the different situations that may have been experienced in the process. Case 1 (Figure 15) was different, not only due to its very high emission values, as discussed above, but also due to its own chronological development demonstrating the process experienced by many households in the city. In the period prior to confinement (February), a level around 800–900 ppb TVOCs with cyclical oscillations associated with use was observed. However, the deterioration of the health situation prior to the declaration of the pandemic progressively led to an increased presence in the home starting the weekend of 22 February. At this time, emissions started to trend upward due to the increase in the intensity of use of the home, as well as a possible increase in cleaning tasks. A milestone was noted on 9 March, when school activities were suspended and a preventive confinement—at least partial for the family unit—created the situation that occurred during the entire state of alarm. In this situation, the typical values of emissions were above 10,000 ppb with very wide variations in concentration depending on the emission and ventilation cycles. In this particular case (CS1), the concentration fluctuations during the confinement peak may have been associated with the intensification of the working activity of a member of the family unit included in the group of “essential workers” who worked during different times during the confinement. The presence of a family member who leaves the home periodically necessitated the intensification of cleaning procedures, personnel, goods, and clothing (more frequent washing of clothes), as well as the incorporation of disinfectant agents more frequently than would be expected in a confinement situation. Int. J. Environ. Res. Public Health 2020,17, 7183 23 of 33 To this, the presence of the smoker inhabitant and the use of air fresheners and odor neutralizers (electric diffusers and aerosols) must be added. Figure 15. TVOC concentration temporal profiles (ppb) for previous operation (blue) and pandemic declaration-time (red) for Cases (a)1(y-axis logarithmic), (b) 2, (c) 3, and (d) 4. All these factors had a strong impact on a relatively airtight dwelling—thus, with lower background dilution—siding with the fact that it was the home with the smallest volume per occupant, which allowed concentrations to rise more than in other cases with a greater dilution reserve. Although, conversely, the CO 2 level was inside the acceptable threshold to WHO standards. This aspect can highlight the need to review these figures, as well as the limitation in the isolated use of CO 2 as an index of environmental quality. In most of the situations, the use of other complementary parameters will be needed. In several situations, the carbon dioxide levels within what has traditionally been considered acceptable may mask the presence of other pollutants and risks. This situation has been previously identified [115–119] and should be subject to future considerations. It is of interest to observe the hourly profile of the concentrations and the great dependence on the activity cycles of the dwelling, as well as those linked to the voluntary ventilation cycles. Taking case 1 as an example, it is possible to identify the impact on the environmental quality of the activity profiles. In case 1, the emission profile was strongly influenced by the hours of rest, where a continuous base level was maintained, but a slight nocturnal dissipation with actuation of the infiltration, until a morning spike, was linked to the beginning of housing activity, with two valleys, preceding the peaks during the central activity time of the day. The shift in values and the alteration of shape of the distribution is clearly evident in Figure 16. Int. J. Environ. Res. Public Health 2020,17, 7183 24 of 33 Figure 16. Hourly median profile of TVOC (ppb) and CO2(ppm) for CS1. Figure 17 shows the process to which the dwellings were subjected during the lockdown. Figure 17. At home lockdown cycle. Int. J. Environ. Res. Public Health 2020,17, 7183 25 of 33 Future confinement events can be expected related to subsequent waves of SARS-CoV-2—or variants— [ 120 – 123 ] as well as new pandemic crises [ 124 ]. These stay-at-home orders could affect the entire population again or be directed to specific sectors of the population, such as the elderly, immunocompromised, or children [ 121 , 125 , 126 ], i.e., the more vulnerable parts of the population. The rising in home-schooling can also increase the time spent at-home or in semi-confinement and, therefore, extend the possible risks even without the application of confinement orders. Thus, there is a need for residential IAQ improvement to answer to the new social challenges. The climatology of the 1st wave period in the city allowed a restrained use of heating, thus contributing to keeping city heating system emissions moderate. Along with the reduction of road traffic, this resulted in a much better urban atmosphere than is typical. As expected, due to climatic conditions the presence of tropospheric ozone (O3) in this same period had no meaningful impacts. While the primary concerns regarding city outdoor pollutants are typically NO 2 /x and particulate matter, and ozone during the warm season [ 56 ], city outdoor conditions were somehow acceptable prior to the lockdown and were good when the lockdown was at full force. It, therefore, seems possible to allocate a low influence from outdoor sources in the evolution of the home indoor situation for comparison purposes, with the increase in internal emissions and the ventilation profiles being primarily responsible for the changes. Fine particulate matter, represented as PM 2.5 , showed evidence of comorbidity in SARS-2 cases, with a relationship between the exposure to the pollutants and the exacerbation of symptoms and increased risk of mortality in the case of contamination of COVID-19 viruses [127,128]. There are risks linked to future confinement episodes in more extreme weather periods. During the coldest climate scenarios—colder than those suffered in this first COVID-19 emergency—the increased use of heating systems, with the associated boiler pollutant emissions, can noticeably increase the presence of pollutants in the city atmosphere. This situation can further aggravate the presence of high atmospheric pressure events during the winter, thus, preventing the atmospheric dilution of contaminants. In future confinements, the residential buildings may have poorer outdoor air to achieve ventilation, resulting in a lower capacity to provide adequate indoor dilution of an often already compromised indoor environment. Cold periods can also have an impact in low performance buildings and in low-income population groups, which very often coincide. The need for heat conservation and to save energy, particularly in homes affected by energy poverty, can result in an even more significant restriction of voluntary ventilation—the only system available to these houses. 5. Conclusions The need for inhabitants to use heating temperatures above those usually recommended by standards was identified during this period. This corresponded to the need to compensate for low radiant temperatures due to the limited performance of their thermal building envelopes. In most cases, the thermal systems of the homes were unable to maintain stable indoor thermal conditions, with a high dependence on outdoor conditions, which indicates either inadequate sizing/efficiency or underperformance of the building materials to control heat loses (this is of particular note in legacy buildings). This usually results in a limitation on the voluntary ventilation. This situation can be further aggravated in future confinement scenarios (SARS-CoV-2/x next waves) during colder periods—mid winter cold waves—where situations of reduced ventilation to conserve heat may coincide with unsuitable outdoor air quality scenarios, even under situations of a radical reduction in road traffic. The increase in the intensity of housing use generates high to very high pollutant rising on indoor emissions, which cannot always be counteracted by ventilation in the dwellings, either due to its reduced effectiveness or due to voluntary limitations. The increase in the domestic use of cleaning and disinfection products during the COVID-19 crisis also have contributed to the increase in VOCs and other chemicals in the indoor environment, just as the generalized imposition of teleworking may have increased the sources of emissions (increase in Int. J. Environ. Res. Public Health 2020,17, 7183 32 of 33 93. Sangiorgi, G.; Ferrero, L.; Ferrini, B.S.; Porto, C.L.; Perrone, M.G.; Zangrando, R.; Gambaro, A.; Lazzati, Z.; Bolzacchini, E. Indoor airborne particle sources and semi-volatile partitioning effect of outdoor fine PM in offices. Atmos. Environ. 2013,65, 205–214. [CrossRef] 94. Cao, J.J.; Huang, H.; Lee, S.C.; Chow, J.C.; Zou, C.W.; Ho, K.F.; Watson, J.G. Indoor/outdoor relationships for organic and elemental carbon in PM2.5 at residential homes in Guangzhou, China. Aerosol Air Qual. Res. 2012,12, 902–910. [CrossRef] 95. Höllbacher, E.; Ters, T.; Rieder-Gradinger, C.; Srebotnik, E. Emissions of indoor air pollutants from six user scenarios in a model room. Atmos. Environ. 2017,150, 389–394. [CrossRef] 96. Kim, S.; Hong, S.H.; Bong, C.K.; Cho, M.H. Characterization of air freshener emission: The potential health effects. J. Toxicol. Sci. 2015,40, 535–550. [CrossRef] [PubMed] 97. Nørgaard, A.W.; Kudal, J.D.; Kofoed-Sørensen, V.; Koponen, I.K.; Wolkoff, P. Ozone-initiated VOC and particle emissions from a cleaning agent and an air freshener: Risk assessment of acute airway effects. Environ. Int. 2014,68, 209–218. [CrossRef] [PubMed] 98. Wang, F.; Meng, D.; Li, X.; Tan, J. Indoor-outdoor relationships of PM2.5 in four residential dwellings in winter in the Yangtze River Delta, China. Environ. Pollut. 2016,215, 280–289. [CrossRef] 99. He, C.; Morawska, L.; Hitchins, J.; Gilbert, D. Contribution from indoor sources to particle number and mass concentrations in residential houses. Atmos. Environ. 2004,38, 3405–3415. [CrossRef] 100. Elbayoumi, M.; Ramli, N.A.; Md, N.; Yusof, F.F. Spatial and temporal variations in particulate matter concentrations in twelve schools environment in urban and overpopulated camps landscape. Build. Environ. 2015,90, 157–167. [CrossRef] 101. WHO. World Health Organization-WHO Guidelines for Indoor air Quality: Selected Pollutants; WHO: Geneva, Switzerland, 2010. 102. Danzon, M.A.; F.X.Van Leeuwen, R.; Krzyzanowski, M. Air Quality Guidelines: For Europe; World Health Organization Regional Publications—European Series; WHO: Geneva, Switzerland, 2001. 103. Zhang, Y. Indoor Air Quality Engineering; CRC Press: Boca Raton, FL, USA, 2004. 104. Chi, C.; Chen, W.; Guo, M.; Weng, M.; Yan, G.; Shen, X. Law and features of TVOC and Formaldehyde pollution in urban indoor air. Atmos. Environ. 2016,132, 85–90. [CrossRef] 105. Meciarova, L.; Vilcekova, S.; Budaiova, Z. Organic compounds in residential buildings. In Proceedings of the Healthy Buildings Europe 2015, HB 2015—Conference Proceedings, Boulder, Colorado, USA, 19–22 July 2015. 106. Meˇciarov á , L’.; Vilˇcekov á , S.; Kr í dlov á Burdov á , E.; Kisel á k, J. Factors Effecting the Total Volatile Organic Compound (TVOC) Concentrations in Slovak Households. Int. J. Environ. Res. Public Health 2017 ,14, 1443. [CrossRef] [PubMed] 107. Lundgren, B. Sampling Strategies for Volatile Organic Compounds in Indoor Air; European Commission 1995. Available online: http://www.inive.org/medias/ECA/ECA_Report14.pdf (accessed on 28 September 2020). 108. Larsen, L. Environmental Health Sourcebook; Omnigraphics Publishing: Detroit, MI, USA, 2010. 109. Nathanson, T. Health Canada Sant é Canada Indoor Air Quality in Office Buildings: A Technical Guide; Federal-Provincial Advisory Committee on Environmental and Occupational Health, Minister of National Health and Welfare, Canada. 1995. Available online: https://irp-cdn.multiscreensite.com/562d25c6/files/uploaded/ Canada_Indoor%20Air%20Quality%20in%20Office%20Buildings_A%20Technical%20Guide_1995.pdf (accessed on 28 September 2020). 110. Epa, U. An Introduction to Indoor Air Quality (IAQ). In Indoor Air Last Updat; United States Environmental Protection Agency: Washington, DC, USA, 2011; pp. 3–5. 111. H é roux, M.E.; Clark, N.; Ryswyk, K.V.; Mallick, R.; Gilbert, N.L.; Harrison, I.; Rispler, K.; Wang, D.; Anastassopoulos, A.; Guay, M.; et al. Predictors of indoor air concentrations in smoking and non-smoking residences. Int. J. Environ. Res. Public Health 2010,7, 3080–3099. [CrossRef] [PubMed] 112. Kaunelien ˙ e, V.; Meišutoviˇc-Akhtarieva, M.; Martuzeviˇcius, D. A Review of the Impacts of Tobacco Heating System on Indoor Air Quality Versus Conventional Pollution Sources; Chemosphere 206; Elsevier Ltd.: Amsterdam, The Netherlands, 2018; pp. 568–578. 113. Steinemann, A. Ten questions concerning air fresheners and indoor built environments. Build. Environ. 2017 , 111, 279–284. [CrossRef] Int. J. Environ. Res. Public Health 2020,17, 7183 33 of 33 114. Liu, X.; Mason, M.A.; Krebs, K.A. VOC emissions from an air freshener in the indoor environment. In Proceedings of the Indoor Air Quality Problems and Engineering Solutions, RTP, United States Environmental Protection Agency, Durham, NC, USA, 21–23 July 2003. 115. Fern á ndez-Agüera, J.; Dom í nguez-Amarillo, S.; Alonso, C.; Mart í n-Consuegra, F. Thermal comfort and indoor air quality in low-income housing in Spain: The influence of airtightness and occupant behaviour. Energy Build. 2019,199, 102–114. [CrossRef] 116. Kapalo, P.; Vilcekova, S.; Domnita, F. Determining the Ventilation Rate inside an Apartment House on the Basis of Measured Carbon Dioxide Concentrations—Case Study. Environ. Eng. 2017. [CrossRef] 117. Guyot, G.; Sherman, M.H.; Walker, I.S. Smart ventilation energy and indoor air quality performance in residential buildings: A review. Energy Build. 2018,165, 416–430. [CrossRef] 118. Moreno-Rangel, A.; Sharpe, T.; Musau, F.; McGill, G. Field evaluation of a low-cost indoor air quality monitor to quantify exposure to pollutants in residential environments. J. Sens. Sens. Syst. 2018 ,7, 373–388. [CrossRef] 119. Shin, M.S.; Rhee, K.N.; Lee, E.T.; Jung, G.J. Performance evaluation of CO 2 -based ventilation control to reduce CO 2 concentration and condensation risk in residential buildings. Build. Environ. 2018 ,142, 451–463. [CrossRef] 120. Campano-Laborda, M.; Dom í nguez-Amarillo, Á .; Fern á ndez-Agüera, S.; Acosta, J.I. Indoor comfort and symptomatology in non-university educational buildings: Occupants’ perception. Atmosphere (Basel) 2020 , 11, 357. [CrossRef] 121. L ó pez, L.; Rod ó , X. The end of social confinement and COVID-19 re-emergence risk. Nat. Hum. Behav. 2020 , 4, 746–755. [CrossRef] 122. Xu, S.; Li, Y. Beware of the second wave of COVID-19. Lancet 2020,395, 1321–1322. [CrossRef] 123. Adebowale, V.; Alderson, D.; Burn, W.; Dixon, J.; Godlee, F.; Goddard, A.; Griffin, I.; Henderson, K.; Horton, R.; Marshall, R.; et al. Covid-19: Call for a rapid forward looking review of the UK’s preparedness for a second wave -an open letter to the leaders of all UK political parties. BMJ 2020 ,369. [CrossRef] [PubMed] 124. Ali, I. COVID-19: Are we ready for the second wave? Dis. Med. Public Health Prep. 2020 , 1–3. [CrossRef] [PubMed] 125. Roaf, S.; Crichton, D.; Nicol, F.; Rudge, J.; Kovats, S. Adapting Buildings and Cities for Climate Change; Routledge: Abingdon-on-Thames, UK, 2010. 126. Ishii, T. Measures Against the Spread of COVID-19: Simultaneous International Action and Lockdown Postponement. SSRN Electron. J. 2020. [CrossRef] 127. Alvarez, F.; Argente, D.; Lippi, F. A Simple Planning Problem for COVID-19 Lockdown. SSRN Electron. J. 2020. [CrossRef] 128. Liang, D.; Shi, L.; Zhao, J.; Liu, P.; Schwartz, J.; Gao, S.; Sarnat, J.; Liu, Y.; Ebelt, S.; Scovronick, N.; et al. Urban Air Pollution May Enhance COVID-19 Case-Fatality and Mortality Rates in the United States. medRxiv 2020. [CrossRef] © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).