National surveys showed the prevalence of marijuana vaping among US adolescents increased from 2017 to 2019 . Vaping among Grade 12 students in the last 30 days increased from 4.9% in 2017 to 7.5% in 2018, reaching 14% in 2019 . Cecinato et al. measured THC in four homes in Rome, Italy, reporting that the indoor THC concentration in two homes was 6.6 ng/m3 indoors and 1.1 ng/m3 outdoors. Indoor concentrations of THC and cannabidiol often exceeded those measured at outdoor fixed air monitoring stations, indicating the sources were indoors. Chu et al. conducted a two-stage probability telephone survey of 2,812 respondents living in multi-unit housing in Ontario, Canada, reporting that 7.5% reported being exposed involuntarily to secondhand cannabis smoke. The prevalence of involuntary exposure to cannabis smoke in multi-unit buildings was similar to that from secondhand tobacco smoke. Moir et al. compared a large number of toxic air pollutants produced by mainstream and side stream marijuana and tobacco smoke, using a smoking machine to produce the smoke. For both marijuana and tobacco cigarettes, the mass of the pollutants measured in side stream smoke was much greater than the mass of the pollutants measured in mainstream smoke for the majority of air pollutants. The amount of benzene, a known human carcinogen,cannabis grow system measured in secondhand marijuana smoke was about the same as that measured in secondhand tobacco smoke . These investigators also compared 30 poly cyclic aromatic hydrocarbons measured in marijuana and tobacco smoke. They found that side stream marijuana smoke contained about the same amount of the human carcinogen benzopyrene as side stream tobacco smoke .
They also found that marijuana side stream smoke contained about 1.5 times the amount of benzo anthracene, another probable human carcinogen, as did tobacco smoke. In both mainstream and side stream marijuana smoke, they reported the presence of many similar known carcinogens and other chemicals implicated in respiratory diseases. Graves et al. compared tobacco and marijuana smoke particles and found them quantitatively similar in volatility, shape, density, and number concentration, with differences in particle size and chemical composition. Their study detected 4350 different compounds in tobacco smoke and 2575 different compounds in marijuana smoke, with 231 compounds common to both tobacco and marijuana smoke. Of these, 173 different tobacco smoke compounds and 110 marijuana smoke compounds were known to cause adverse health effects through carcinogenic, mutagenic, teratogenic, or other toxic mechanisms. Ni et al. reviewed dozens of health studies on PM2.5 related to tobacco smoking, concluding that indoor PM2.5 from cigarette smoking is closely correlated with chronic lung disease. Due to the small size of these particles, they can go deep to the distal airways and deposit in alveolar regions, doing serious harm to the human respiratory system. They carry with them PAHs and many other toxic compounds. Although there are few studies of the health effects of marijuana aerosols, Wang et al. reported that 1-min of exposure to secondhand cannabis smoke can impair vascular endothelial function in rats. In the US, marijuana is most often smoked indoors in homes , but relatively few measurements have been made of indoor air pollution from cannabis use in residences. Californians are prohibited from consuming legal cannabis in “any public place or area” or in “any location where tobacco smoking is prohibited,” although legal cannabis can be consumed in private residences or in structures located on the grounds of a private residence . Other residents of a home may find the odor from marijuana smoking objectionable, so the smoker may confine his or her smoking activity to a room with a closed door.
To provide data on the concentrations and emissions produced by cannabis use indoors in a home, we conducted 60 controlled experiments in the spare bedroom of an occupied residence. Twenty-four experiments were conducted on pre-rolled cannabis joints, and 9 experiments each were conducted on bongs, glass pipes, vaping pens, and tobacco cigarettes. These measurements of PM2.5 from secondhand cannabis smoke were compared with PM2.5 from secondhand tobacco smoke from Marlboro cigarettes , the most popular cigarette brand in the US . Zhao et al. conducted cannabis experiments in a car using joints, bongs, glass pipes, and vaping pens as sources, developing gravimetric calibration factors for four different methods of smoking marijuana. These calibration factors were used for the same sources in the present study. Wallace et al. measured secondhand exposure to PM2.5 from vaping marijuana in two different homes. To our knowledge, these efforts are the first systematic studies measuring PM2.5 mass concentrations, source strengths, emission rates, and decay rates from secondhand cannabis smoke indoors in residences. A habitual user of cannabis and tobacco, who consumes cannabis in multiple ways, was recruited to help generate secondhand cannabis smoke. The study protocol was accepted by the participant, and a signed consent form was obtained before the experiments.No individual other than the participant was involved in the smoking or vaping activities, and no persons were present in the room during the air pollutant decay periods. The main focus of this research was on comparing the emission rates produced by different methods of smoking or vaping cannabis sources, not on the health impact on human subjects. The study protocol was approved by the Institutional Review Board at Stanford University. This study was supported by a grant awarded to Stanford University to study secondhand exposure to marijuana: Agreement #28IR-0062 sponsored by the University of California Office of the President; Tobacco Related-Disease Research Program . The cannabis materials used in this study were provided by the participant. The 60 controlled experiments measuring fine particle mass concentrations were carried out in an occupied residence in Redwood City, CA, on 24 dates between April 16 and November 25, 2019. All the experiments were conducted in a 43 m3 spare bedroom that was set off from the rest of the house. This room had one window and one door, both of which were closed prior to the start of each experiment. The 60 experiments were conducted on 23 different dates, with one experiment conducted on each of three dates, two experiments conducted on five dates, three experiments on 13 dates, and four experiments on two dates. On dates with more than one experiment, the room’s window and door were opened before each experiment to air out the room, and the home’s front door, backdoor, and a kitchen window also were temporarily opened. All experiments took place during the daytime hours, each lasting about 2–1/2 h. This provided sufficient time for mixing in the room and allowed for estimation of the PM2.5 decay rate. Prior to starting each experiment, the monitors were operated for at least 10 min to measure the background PM2.5 concentrations in the room. The background concentrations were relatively small and were subtracted prior to analyzing the PM2.5 concentration data. The heating and air conditioning system of the home was turned off before and during all experiments, and the home’s exterior doors and windows were closed. We compared the PM2.5 emissions produced by four different methods of consuming marijuana grow system – joint, bong, glass pipe, and vaping pen – with the emissions from Marlboro tobacco cigarettes purchased in California in 2019. The first three cannabis consumption methods use combustion to produce PM2.5, while the vaping pen uses a heated coil to vaporize cannabis liquid without combustion. We used TSI AM510 SidePak™ laser photometers with the individual calibration factors of each monitor based on gravimetric measurements .
The calibrated SidePak mass measurements were found to agree well with measurements by the piezoelectric micro-balance , an instrument that measures real-time mass concentrations directly. All the cannabis joints, bongs, buds, and vaping supplies used in this study were commercially available and were purchased from four state licensed stores in three California towns in 2019: San Jose, Palm Desert, and Cathedral City. The 24 pre-rolled marijuana joints used in this study consisted of 9 different name brands that are widely available in California. A factory label that came with each joint listed its CBD and THC content. The CBD content of the 24 joints ranged from 0% to 1.5%, and the listed THC content ranged from 8.55% to 27.6%, with a mean of 17.7%. We used a laboratory scale to measure 0.3 g of cannabis buds into the bowls of the bong and the glass pipe, and the two types of cannabis buds used were “Mirage” and “Blueberry Muffin” . The electronic vaping pen was manufactured by AbsoluteXtracts , and we attached two different vaping cartridges to the pen: a Care by Design 18:1 cartridge and a Care by Design 2:1 cartridge . In each experiment, we used at least 2 AM510 SidePak monitors with their individual calibration factors based on the gravimetric filter measurements obtained by Zhao et al. . Each SidePak was equipped with a physical 2.5 μm size impactor supplied by the manufacturer, and the data logging times were set to 1.0 min. Before starting each set of experiments, the grease on the monitor’s particle size impactor was replaced, and the monitor was zeroed using a precision zero filter supplied by the manufacturer. Periodically we measured the flow rate of each monitor using a Gillibrator Primary Flow Calibrator , verifying that it was within 5% of the 1.7 L/min flow rate specified by the manufacturer. We also used precision digital clocks synchronized with the atomic clock in Boulder, CO, to verify that the data logging times of each monitor were within ± 3 s of the correct time. Each experiment used a pair of SidePak monitors for redundancy and sometimes a third SidePak monitor as a backup. Each SidePak’s internal calibration factor was set to 1.0, and the proper calibration factors for the monitors and source type, based on the gravimetric measurements of Zhao et al. , were applied subsequently in the data analysis phase. A comparison of the two main SidePak monitors, each using a calibration factor for cannabis vaping based on gravimetric filter measurements, showed good agreement . In these experiments, we also used a pair of TSI 3007 condensation particle counters to measure ultrafine particles greater than 10 nm in diameter at 1-min time intervals. The monitors were placed near the midpoint of the room at a height above the floor of 0.6 m, and a small battery-powered fan with an 11 cm diameter blade was run for the duration of each experiment to assist with air mixing. Immediately after the smoking or vaping ended, the participant exited the room, carefully closing the door behind him. Thus, no one was exposed to secondhand smoke in the room during the decay period of about 100–130 min. A video camera was set up in the room with its lens pointed toward a SidePak monitor’s display screen, sending readings of the measured PM2.5 concentrations to computer screens outside the room. We used a pair of Model T15n electrochemical Carbon Monoxide Measurers™ to measure the CO concentrations in the room produced by releasing CO from a 105 L cylinder containing 10% CO gas in nitrogen . The CO gas was emitted into the room for approximately 6 min prior to the start of each experiment using an adjustable flow rate regulator set to 1 L/min. The resulting CO decay rate was used to estimate the room’s air exchange rate, based on the negative slope of the logarithm of the background-corrected CO concentration. Ferro et al. used Brüel and Kjær Type 1302 photoacoustic sulfur hexafluoride monitors to measure the volume of this same room and its air exchange rate, and our air exchange rate measurements were consistent with their published measurements. Our main objective was to compare the concentrations and emission rates produced by different methods of cannabis smoking and vaping, so it was important to apply the same procedure to each source in these experiments. All the smoking or vaping methods in these experiments followed the 3-Puff Protocol, which consisted of a starting puff at time t = 0, followed by a 2nd puff at t = 60 s, followed by a 3rd puff at t = 120 s . When a joint or a cigarette reached 3.0 min, the participant put it out by dipping the tip in water. This protocol is well-suited to combustion sources, which produce both mainstream and side stream smoke.