Subjects were required to live in the northeastern U.S. and have an infant age 0–6 months at the time of enrollment. No other exclusion criteria were applied. Recruitment began June 2009 and continued until October 2009 and was conducted through distribution of flyers in thirteen maternity clinics, hospitals, and birthing classes throughout the northeastern U.S. Consent was obtained from the places of business where the flyers were posted. Advertisements were posted online on new mother websites and community billboards. In exchange for participation in the phone survey, each participant was offered $20 to be paid after the survey. At the conclusion of the study, all participants were offered results from the study and a brochure on indoor air quality. Subjects who participated in the air monitoring were offered the monitoring results for their home. The Yale Human Subjects Committee approved this study. The five stages of the study design were thus: recruitment; phone survey; indoor air monitoring; data analysis; and information distribution to study participants.

Indoor air monitoring was conducted for carbon monoxide (CO), carbon dioxide (CO₂), particulate matter with aerodynamic diameter ≤0.5 (PM_0.5), total volatile organic compounds (TVOCs), temperature, and humidity. These pollutants were chosen based on previous observed health impacts and compatibility with available monitors [15,20,21,22,23,24,25,26]. Measurements were taken between June 19 and August 9, 2009 in ten homes: three homes in Connecticut, three in New York, two in Vermont, and two in Massachusetts. The air monitor was placed in the nursery approximately two feet above the floor on a horizontal surface (e.g., chair, shelf, table) away from windows and doors, and measurements were taken for 4 to 7 days in each home. For the purpose of this study, the term “nursery” is used to describe the area in the home where the infant spends the majority of his/her time, as identified by the participants.

Monitoring was conducted with an AirAdvice M7100 [20,21]. TVOCs and PM_0.5 are reported in µg/m³ and CO₂ and CO in ppm. This monitor does not measure other forms of PM, such as particulate matter with aerodynamic diameter ≤2.5 µm or ≤10 µm (PM_2.5 or PM₁₀). For measurements that exceeded the CO₂ monitor limit of 2000 ppm, summary measures (e.g., overall average of CO₂ in a home during the monitoring period) were calculated with these values as 2000 ppm, which underestimates the true pollution levels. The pollutant measurements were analyzed to determine the variability and overall levels of each pollutant within the nurseries and among the participants. Variability was assessed by summary statistics, boxplots, analysis of variance, and coefficient of variance (COV), which is the standard deviation divided by the mean ×100%, with larger values indicating higher variability in relation to the mean value.

A telephone survey of 53 study subjects was conducted to assess potential sources of residential indoor air pollution, and to investigate the degree to which the arrival of a new infant coincided with activities that may relate to indoor air pollution (e.g., indoor painting). The 10 subjects who participated in air pollution monitoring also participated in the survey, and were randomly selected. Questions elicited information on the parent characteristics (e.g., mother’s education) and household characteristics (e.g., type of heating fuel). The survey focused on aspects of the home that may impact air quality, such as renovation prior to or shortly after the infant’s arrival, air purifiers, and presence of cockroaches [27,28,29,30,31,32]. Questions pertaining to renovation refer to activities taking place in the last six months of pregnancy to the time of the survey. All surveys were conducted by a single investigator to ensure consistency.

At the conclusion of the study, each participant was offered the results from the study for overall findings and a U.S. Environmental Protection Agency (U.S. EPA) brochure providing general information on indoor air quality [33]. Participants in the monitoring phase of the study were offered monitoring results for their own home.

Monitoring for indoor air quality was conducted in the room in which the infant spends the most time for 10 homes. Table 1 summarizes the pollutant levels, temperature, and humidity in each home. Pollutant levels varied greatly across homes with overall levels for some homes up to 20 times higher than for another home (e.g., TVOCs for Participants D and E). The homes with the highest levels varied by pollutant (Participants F and I for CO, D and I for CO₂, B and C for PM_0.5, and D and F for TVOCs). Those with the highest overall levels were not necessarily those with the highest peak levels. For instance, Participant A’s average PM_0.5 level was lower than that of most other homes (6 of 9), but had the highest hourly value.

Figure 1 provides boxplots of hourly pollutant levels for each pollutant and home, and illustrates the variability of pollutant levels within a given nursery and across homes. Ninety percent of hourly measurements were between 0.37 and 1.93 ppm for CO, 411.9 and 1194 ppm for CO₂, 4.4 and 53.3 µg/m³ for PM_0.5, and 124.0 and 5171 μg/m³ for TVOCs. Figure 1 also provides the COV for each pollutant and home. Participant G had the highest variability, based on the COV, for CO, CO₂, and TVOCs, and the second highest variability for PM_0.5. CO₂ exhibited the lowest average variability in relation to the mean and PM_0.5 the highest. Between participant variability was larger than within participant variability for all pollutants, based on analysis of variance (p-value <0.001). TVOC levels appeared highly variable both across the nurseries and within individual nurseries (Figure 1d), with average levels within a home ranging from 210.8 to 4683 µg/m³ and some measurements exceeding 6000 µg/m³.

We calculated Pearson’s correlations among all pollutant pairs for each home (Table 2). The highest correlations were among TVOC and CO₂ although this relationship varied across homes (average correlation 0.68, range 0.42 to 0.94). Other high correlations were noted for CO and CO₂ (average 0.60, range 0.06 to 0.90). Table 2 also shows the correlations between pollutant levels and weather variables. On average across the 10 homes, pollutants were not correlated with temperature or relative humidity, although one home had a correlation of 0.71 for humidity and CO₂, and another of 0.81 for humidity and PM_0.5.

Results indicate that the daily pattern of exposure to indoor air pollution in nurseries also differs across homes. As an example, Figure 2 shows hourly average levels of TVOC concentrations across time in three homes. Levels in Participant D’s home averaged 4683 μg/m³ and exceeded 6000 μg/m³ at some point each day. Levels for Participant E followed a different pattern, averaging 210.8 μg/m³ and below 500 μg/m³ for most of the measurement period, then rising sharply to an hourly average >1,300 μg/m³ during the final day.

TVOCs in Participant G’s home averaged 1343 μg/m³ and varied widely with full-day (24-h) daily averages ranging from 832 to 1638 μg/m³ and hourly averages >5000 μg/m³ on two days. Supplemental Figures 1–4 provide similar graphs for all four pollutants and homes.

Indoor air quality in residences is not regulated; however for context, we compared the measured levels to several air quality guidelines and standards including health-based regulations for outdoor air, regulations for indoor air in occupational settings, and indoor air quality guidelines. Table 3 provides air quality guidelines developed by AirAdvice using information from the U.S. EPA; the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE); World Health Organization (WHO); Leadership in Environmental and Ecological Design (LEED); Indoor Air Quality Association; California Air Resources Board; Health Canada; European Union; U.K. Department of Health; state governments; and scientific experts [20,21]. Table 4 provides the air quality metrics for each pollutant based on the exposure timeframe used in Table 3 (i.e., daily average for PM_0.5, CO₂, and TVOCs; maximum 1-h and 8-h average for CO) for each participant and each day (Table 4). We also compared measurements to air quality guidelines developed by the Occupational Safety and Health Administration (OSHA), the American Conference of Governmental Industrial Hygienists (ACGIH), and the National Institute for Occupational Safety and Health (NIOSH) for indoor air in occupational settings. Other comparisons include guidelines from the Consumer Product Safety Commission, residential air quality guidelines from Health Canada, and U.S. EPA health-based regulations and WHO guidelines for outdoor air.

The daily maximum 1-h CO level for each day and nursery are provided in Table 4a. CO levels for all nurseries met the guidelines in Table 3 with all hourly values <20 ppm and 8-hour values <5 ppm. Measurements of CO met the Consumer Product Safety Commission recommended maximum levels of 15 ppm for 1 h and 25 ppm for 8 h, and the Health Canada recommendations for residential air quality at 25 ppm for 1 h and 10 ppm for 24 h [34]. The measured levels also were well within standards and guidelines for occupational settings: the OSHA permissible exposure limit (PEL) for general industry of 50 ppm for an 8-h exposure [35], the NIOSH recommended exposure limit (REL) of 35 ppm 8-h average based on cardiovascular response [36], and ACGIH threshold limit value of 25 ppm 8-h average based on risk of elevated carboxyhemoglobin levels [37]. Measured CO was below EPA’s National Ambient Air Quality Standards of 9 ppm for 8 h and 35 ppm for 1 h [38] and the WHO’s guidelines for ambient air for 10 ppm for 8-h and 25 ppm for 1 h [39].

Table 4b compares daily CO₂ levels for a given nursery to air quality guidelines. For four nurseries at least one daily CO₂ level exceeded the guideline for action needed, >999 ppm as per Table 3, and for one of these homes, that level was reached three out of four days measured. An additional household had one daily average at a level for recommended action for sensitive individuals. For four of the homes, CO₂ levels exceeded ASHRAE recommendations that indoor levels be ≤1000 ppm for schools and ≤800 ppm for offices [40]. Recommendations for occupational settings are much higher at NIOSH guidelines of 5000 ppm for an 8-h exposure and 30,000 ppm for any 10-min period, OSHA PEL of 5000 ppm for an 8-h period [35]. Levels were also well below ACGIH guidelines of 5000 ppm for an 8-h exposure and 30,000 ppm for a short-term exposure limit [41]. Based on AirAdvice criteria, all homes had PM_0.5 and TVOC levels above the threshold for action recommended for sensitive groups (Table 4c,d). The levels for action necessary were exceeded by three homes for PM_0.5 and half of the homes for TVOCs.

Supplemental Table 1 describes the participants in the telephone survey (N = 53). Most respondents were mothers, with average age of 31.3 years (range 19–43 years) and infants with an average age of 97 days (range 6–192 days). No participants were pregnant at the time of the survey. Participants were predominately white (85%). The subjects’ educational levels were high with all mothers having completed at least one year of college, and almost 19% with 20 or more years of school. More than 79% of participant households had annual incomes exceeding $50,000. Of the 25% of homes with older siblings, 15% of those siblings had asthma and 38% had allergies. The infants spent most of their time in the home, with 9.4% in daycare (average 36 h/week).

Table 5 describes housing characteristics for the study participants. Most houses were older (average year built was around the early 1940s, range 1810–2005). The most common home heating sources were oil, gas, and electric, with only 5.7% for wood, and some homes having multiple heating sources. Other potential sources of indoor air pollution identified by 20% or more of participants were fireplaces (21%), gas stoves (62%), attached garage (28%), mice (32%), and indoor pesticides (21%). Of the 51% with pets, 44% had cats and 63% had dogs. Aromatic candles were in the nursery for 15.1% of the nurseries. Of the candle users, 38% kept the candles unlit. Smoking took place in 3.8% of the homes. Survey participants estimated the distance between the home and the nearest road at an average of 63 feet, with 53% less than 25 feet away.

Table 6 describes renovation characteristics for the study participants. Of the 53 participants, 66% had remodeled the nursery at some point during the last six months of pregnancy to the time of the survey, with fewer (42%) having remodeled another portion of the house. Of those that remodeled the nursery, 11% of participants responded that renovations were complete before the infant’s arrival, on average 52 days prior to birth (range 24–92 days). Painting had taken place in the nursery for 49% of the participants, with interior walls painted more often than furniture, floor, or trim. Of those who painted, 46% reported using low-VOC paint. Area rugs were added to 28% of nurseries, 73% of which were new. Those who renovated the nursery did not differ from those who did not by mother’s demographics (age, race, or education) or the age of the home age. Similarly, those who painted the nursery and those who did not were similar on these factors (results not shown).

Several general trends were observed in the survey data, although differences are not statistically significant and sample size is limited. Those in rural environments were more likely to renovate (89%) or paint (78%) the nursery than those in urban environments (6.7% renovated, 44% painted). Of those with household annual income <$40,000, 71% renovated the nursery and 75% lived in rural environments, compared to 55% and 0% of those with annual income >$100,000, respectively. A similar fraction of those with income <$40,000 painted the nursery (57%) as those with incomes >$100,000 (50%); however, painters in the lower income group were more likely to select low-VOC paint (75%) than painters in the higher income group (40%). Those with older children with asthma or allergies were less likely to have potential indoor air pollution sources than those without older children with these health concerns for: wallpaper installation in the nursery (0% vs. 15%), painting of the nursery (40% vs. 50%), use of indoor pesticides (0% vs. 23%), smoking in the home (0% vs. 4.2%), and room deodorizers in the nursery (0% vs. 8.3%). Those with older children with asthma or allergies were more likely to have air purifiers in the nursery (20% vs. 13%) and less likely to have pets (40% vs. 52%).