Indoor PM2.5 Mortality in China when Outdoor Air Meets 2021 WHO AQG

13 The World Health Organization (WHO) Air Quality Guidelines (AQG) 2021 for PM 2.5 is tightened to be 5 14 μg/m 3 . We firstly estimated deaths attributable to human exposure to PM 2.5 (DAHP) to be 455 thousand (372- 15 527) in urban China in 2019, of which indoor sources contributed 253 thousand (207-294) deaths. The 16 economic losses related to PM 2.5 from indoor sources were 0.98 trillion (0.80-1.14) RMB, accounting for 17 56% of the total economic losses. We then further projected the DAHP at 328 thousand (260-392) when the 18 outdoor PM 2.5 concentration is 5 μg/m 3 , while PM 2.5 from indoor sources still causes 297 thousand (235-355) 19 deaths and 1.27 trillion (1.00, 1.51) in economic losses each year. There are significant health hazards and 20 economic losses caused by indoor PM 2.5 , even the outdoor air is clean enough. The formulation and 21 implementation of more air pollution policies are therefore in urgent need to control indoor sources of PM 2.5 .


27
PM2.5 pollution is a global concern. The World Health Organization (WHO) issued a new set of guidelines 28 on air pollution on Sept 22, 2021, tightening the PM2.5 Air Quality Guidelines (AQG) from 10 μg/m 3 to 5 29 μg/m 31 . Particularly, the guidelines stated that the same AQG should be applied to indoor environments.

30
There are important sources of PM2.5 indoors. WHO has been calling for attention to household air pollution 31 due to solid fuels and kerosene in open fires for cooking. However, even in households using clean fuels and 32 technologies, typically in urban areas, cooking and smoking produce a large amount of PM2.5 2, 3 . Our latest 33 study separated the contribution of indoor and outdoor sources for human exposure to PM2.5 in urban China, 34 and we figured out that indoor sources contribute over 50% of total PM2.5 exposure for Chinese urban 35 residents in 2019 4 . However, the health effect and corresponding economic losses related to PM2.5 from 36 indoor sources are not clear. In particular, current policies on air pollution control of various countries still 37 aim at reducing outdoor PM2.5 concentrations, including strengthening industrial emission, electrification of 38 road vehicles, adjustment of energy structure 5 . However, in view of the significant contribution of PM2.5 from 39 indoor sources to human exposure, indoor PM2.5 concentrations are likely to remain very high when outdoor 40 PM2.5 meets the AQG, resulting in considerable health effects and economic losses. In this study, we 41 estimated the deaths and economic losses attributable to PM2.5 from indoor and outdoor sources based on our 42 previous modelled human exposure to PM2.5 from indoor and outdoor originated sources in Chinese urban 43 areas in 2019 4 , and we further analyzed the scenario when the outdoor air meets the WHO AQG 2021.

47
The deaths and economic losses attributable to PM2.5 in Chinese urban areas in 2019.  (Fig. 1a). The DAHP related to chronic obstructive pulmonary disease (COPD), lung cancer 54 (LC), lower respiratory infections (LRI) and type 2 diabetes (DM2) were 176 thousand (139-212) deaths in 55 total. DAHP in Chinese urban areas is splited according to sources of PM2.5 (Fig. 1b). The DAHP from indoor 56 sources were estimated to be 253 thousand  in Chinese urban areas in 2019, of which indoor outdoor air meets the WHO AQG 2021. It is essential to control the PM2.5 from indoor sources to reduce the 80 DAHP.

82
The WHO AQG 2021 for PM2.5 is 5 μg/m 3 applied to both outdoor and indoor environments. The 83 concentration of human exposure to PM2.5 was projected to be about 39.3 μg/m 3 when the outdoor air meets 84 the WHO AQG 2021 in Chinese urban areas. To meet the WHO AQG 2021, an additional 34.3 μg/m 3 85 reduction in human exposure to PM2.5 is required by controlling indoor sources. Cooking and smoking are 86 sources of PM2.5 in residences in Chinese urban areas, with the emission rate of PM2.5 reaching up to 10 87 mg/min 2 and 4 mg/min 3 , respectively. More than 90% of households in Chinese urban areas cook at home at 88 least once a day according to our survey 4 . People usually run range hoods to reduce indoor PM2.5 89 concentration during cooking. But the exhaust efficiencies of range hoods are very low in Chinese homes, 90 less than 60% 2 . High exhaust efficiency but low-energy-consumption range hoods and scientific kitchen 91 layouts, structures, and ventilation designs are urgently needed for households in Chinese urban areas to 92 protect people from exposure to PM2.5 emitted during cooking 6 . Maintaining good habits of using range 93 hoods, such as washing them regularly and running for some time after cooking, are also ways to reduce 94 human exposure to PM2.5 from cooking. The smoking ban has been implemented in China for many years.

95
However, there are still 44.9% of adults and 63.2% of adolescents exposed to second-hand smoke 7, 8 . Strict 96 enforcement of banning smoking is also in urgent need to reduce exposure to secondhand smoke in China.

97
In addition, the use of air purifiers can effectively remove PM2.5 indoors, and there has been evidence 98 supporting the health benefits of using air purifiers 9 . However, they are still not cost-benefit effective 10 .

99
Reducing the cost of air purifiers may be beneficial for their promotion, particularly for middle-and low-100 income groups.

102
More than 70% of the people in the world use clean fuels or technologies 11 . This study showed that  Table S13 in Hu and Zhao (2021) 4 . Then, C from smoking (Csmoking) were 135 calculated with the following equation: where CSHS,indoor is the C from indoor source of smoking households in the population. The value of CSHS,indoor 138 (denoted as IEC for smoking households) was also provided in Table S13 in Hu and Zhao (2021) 4 . PSHS is 139 the proportion of the population exposed to second-hand smoke, calculated by where Psmoking is smoking rate, the subscript group represents the population with specific age and gender,

142
and Pnumber,f is the proportion of the population that the family household size is f. The age-, sex-, and 143 provincial-specific smoking rates in 2019 were calculated using the age-, sex-, and provincial-specific  Table S13 in Hu and Zhao (2021) 4 ) and risk curves.

207
Uncertainty analysis. We applied a two-stage Monte Carlo simulation 10, 27 to obtain the mean and 95% 208 confidence interval of DAHP and economic loss. The two stages reflected the intra-population variability 209 distribution of human exposure to PM2.5 and the uncertainty of the risk curve, respectively. We performed 210 2,000 and 1,000 iterations (2,000,000 runs in total) at variability and uncertainty stages, respectively. We 9 316 26. Wang, H.andMullahy, J. (2006). Willingness to pay for reducing fatal risk by improving air

334
Competing Interest Statement

335
The authors declare no competing interests.

338
Correspondence and requests for materials should be addressed to B.Z.