## INTRODUCTION

Electronic cigarettes (e-cigarettes), also known as electronic nicotine delivery systems (ENDS), are believed to be less harmful to health than combustible cigarettes and purported to be a safer alternative for conventional smokers and a tool for smoking cessation1. However, possible public health benefits can only be achieved if e-cigarette users do not switch to or concurrently use combustible cigarettes2. In addition, the long-term effects of e-cigarettes on health and their use as smoking cession aids are inconclusive3. Some studies have found e-cigarette use increased cessation rates4-6, but other studies have found contradictory results7,8. Despite the inconclusive results about the benefits of e-cigarettes, their use has been increasing in recent years in China and around the world9,10. Meanwhile, heated tobacco products (HTPs such as IQOS ‘I quit original smoking’) have recently been reintroduced to the mass market. While HTPs are classified as tobacco products in some countries such as the US, they are marketed as a variation of e-cigarettes in other countries such as Korea. In China, both HTPs and heated liquid e-cigarettes are referred to as ‘dian zi yan’ (i.e. e-cigarettes) by sellers and consumers. A significant proportion of participants in an earlier study among the Chinese did not possess appropriate knowledge about e-cigarettes11 and many even never heard of e-cigarettes12. Therefore, a deeper understanding of the awareness and perception of e-cigarette use at the population level is critical to developing any preventive measures.

Perceptions of risk can affect tobacco use13, similarly, the perceived health benefits and the effectiveness of e-cigarettes in quitting may affect their usage. Awareness of e-cigarettes was found to be most common among smokers and was related to being educated and being younger14. Female and better-educated smokers were more likely to use e-cigarettes15. In earlier studies among adolescents and young adults, the main reasons to use e-cigarettes were curiosity, attractiveness and appeal of flavors and cleanliness, that they were fashionable, had harm reduction potential, and an aid to quit smoking. However, most of the published studies were conducted in high-income countries with limited studies conducted in developing countries. In China, several recent studies have focused on the urban population in China9,10,16,17, with no studies focusing on e-cigarette use perception among adults in households of rural China.

The number of smokers who know and have tried e-cigarettes is growing fast in China. For example, in the International Tobacco Control Policy Evaluation Project (ITC)-China18, the percentage of smokers who knew of e-cigarettes rose from 29% to 60% and those who had tried e-cigarettes rose from 2% to 11%, between 2009 and 2014. This increasing trend in e-cigarette awareness and use may have a negative impact on the overall tobacco control effort in China and underscore the need for public health measures addressing e-cigarette use. Although the use of e-cigarettes has been increasing in China, the governance of e-cigarette use has been loose compared to other countries that have enacted relatively strict regulations for e-cigarette use and supply19. Until November 2019, the State Tobacco Monopoly Administration of China and the State Administration for Market Regulation issued the ‘Notice on Further Protecting Minors from E-cigarettes’ and banned all e-cigarette sales online20. Understanding public perceptions can inform regulation, prevention messaging and other public health actions for e-cigarette use in China and in other low- and middle-income countries. Also, targeting rural populations is unique in that many of the available tobacco control programs and services are enforced in urban areas while their availability in rural settings is scarce, making rural populations vulnerable to receiving wrong information (i.e. misinformation about e-cigarette benefits) and, thus, engaging in risky behaviors18. Also, as part of our ongoing research on family-based smoking cessation promotion and secondhand smoke exposure reduction interventions, we have found that children aged ≤5 years are vulnerable to secondhand smoke exposure in the household. Our exploratory work in urban China made us realize that certain groups of rural populations (i.e. friends of urban e-cigarette users) use e-cigarettes that are provided by their urban friends21. Given the growing e-cigarette market in China and the lack of tobacco control measures in rural China9, we examined, in the present study, the e-cigarette use perceptions among the rural population. Rather than designing a new study, we utilized the sampling framework of an existing study that was conducted among household members of children aged ≤5 years. Therefore, this study examined how smoking status and sociodemographic factors were associated with perceptions of e-cigarette related harms, benefits, and addictiveness, among the adults in households with children in rural China.

## METHODS

### Settings

The study settings were rural communities in Taizhou and Dali. In these two communities, villages were purposively selected, to cover diverse sociocultural aspects of the local population, as recommended by the local collaborators. Taizhou is located in central Zhejiang province, in the eastern part of China. Taizhou is an economically developed region and the GDP per capita of Taizhou was RMB 72912 (Chinese Renminbi, about US$10000) in 201722. The rural population accounted for 37.8% of the total population (about 6 million) of Taizhou22. Four villages were purposively selected for this study, including two in Luqiao district and two in Linhai city in Taizhou. Dali is the autonomous prefecture for the Bai minority, located in the western part of China. The GDP per capita in 2017 of Dali was RMB 29846 (about US$4500)23. The four villages from Jianchuan country in Dali were purposively selected for this study. The total population of Jianchuan was over 18 million in 2016 and the rural population accounted for 27% of the total population24.

### Sample

The study population consisted of household members of children aged ≤5 years in the selected community. To be eligible, all households needed to have at least an adult (i.e. aged ≥18 years) smoker (who smoked one or more cigarettes daily for the past 30 days) who resided in the same house with the child and was able to communicate in Mandarin Chinese. All the eligible households were approached to participate in the study. These households were part of another large ongoing study addressing secondhand smoke exposure among children. One smoker and the main caregiver of the child were invited to participate in the survey. If a household contained more than one smoker, then the smoker who spent more time with the child was recruited. The study was approved by the ethical review board at Duke Kunshan University (IRB No: 2016ABDU003). All participants consented to participate in our survey. The data used and analyzed in this study are available from the corresponding author on reasonable request.

### Procedure

Local community health workers (CHWs) or village doctors screened households, following the child health records available in the community health center, and invited eligible households to participate in the study. Trained interviewers, accompanied by CHWs, visited households that agreed to participate in the study during an agreed time. The interviewers described the purpose and details of the study and sought informed consent from the participants. After confirming the subjects’ informed consent, interviewers completed the survey interview.

### Data collection

The study was conducted in March 2018 in Taizhou and June 2018 in Dali, separately. Students from local colleges/universities were recruited as interviewers to collect data. Interviewers received 3 hours training to learn the necessary communication and interviewing skills. The researcher took the interviewers through each questionnaire question and explained thoroughly the information required. Practice interview sessions were organized to allow interviewers to learn the skills.

A structured Chinese questionnaire, developed with reference to the available literature, was used for data collection and was pilot tested among ten potential subjects before finalization. The interviewers completed the survey via home visits, accompanied by a local CHW/village doctor. After giving informed consent, each household answered a questionnaire containing two parts: Part I for smokers and Part II for caregivers. If the smoker was the same person as the caregiver, then he/she only needed to answer the questions about the children’s health in Part II. The questionnaire was: 1) filled in by either the subjects themselves or 2) the subjects responded to the questions orally and the interviewers wrote down the answers on their behalf.

### Measures

The two-part questionnaire (Part I for smokers and Part II for caregivers) included the following domains.

#### Sociodemographic characteristics

These included region, type of participants (smoker or caregiver), gender, age, educational level, ethnicity, marital status, and occupation (Table 1).

##### Table 1

Demographic characteristics of participants by smoking status, rural areas of Taizhou and Dali, China 2018 (N=1211)*

DemographicsOverall n (%)Smoker n (%)Nonsmoker n (%)p
Total1211668543
Region (n=1211)0.0001
Taizhou891 (74)462 (69)429 (79)
Dali320 (26)206 (31)114 (21)
Age (years) (n=1141)0.7499
≤34413 (36)234 (36)179 (37)
35–54385 (34)227 (35)158 (32)
≥55343 (30)194 (29)149 (31)
Gender (n=1168)<0.0001
Male676 (58)659 (99)17 (3)
Female492 (42)6 (1)486 (97)
Ethnicity (n=1165)<0.0001
Chinese Han903 (78)474 (72)429 (85)
Minority262 (22)188 (28)74 (15)
Marital status (n=1165)0.0622
Married1143 (98)652 (98)491 (97)
Single12 (1)7 (1)5 (1)
Divorced or widowed10 (1)2 (1)8 (2)
Occupation (n=1164)<0.0001
Farmer397 (34)233 (35)164 (33)
Self-employed182 (16)138 (21)44 (9)
Unemployed206 (18)41 (6)165 (33)
Worker247 (21)177 (27)70 (14)
Other132 (11)74 (11)58 (11)
Education level (n=1157)<0.0001
Primary school or below335 (29)152 (23)183 (36)
Middle school476 (41)307 (47)169 (34)
High school or above346 (30)197 (30)149 (30)

* N may not equal to 1211 for some variables due to missing values.

#### Knowledge about smoking and secondhand smoke (SHS) exposure

The first four items (Table 2) were measured using a 5-point Likert scale with response categories ranging from 1 (minimum risk) to 5 (maximum risk). The last four items were measured using a 5-point Likert scale with response categories ranging from 1 (totally disagree) to 5 (totally agree) with those who responded ‘I don't know’ coded as zero. Higher scores indicated a higher perceived risk on each item. The mean total score for the eight items was calculated to evaluate overall smoking-related risk perception. Smoking-related risk perception combines perception about dangers of smoking and perception about dangers of SHS.

##### Table 2

Smoking and secondhand smoke (SHS) exposure related knowledge among participants by smoking status, rural areas of Taizhou and Dali, China 2018 (N=1211)

VariablesOverall mean±SDSmoker mean±SDNon-smoker mean±SDp*
Smoking increases the risk of lung cancer4.12±1.134.07±1.164.19±1.090.0746
Smoking increases the risk of COPD4.05±1.134.00±1.164.12±1.080.0894
Smoking increases the risk of heart diseases3.74±1.273.65±1.343.85±1.180.0069
Concerned about the harmful effects of SHS exposure to children’s health4.29±1.064.19±1.134.41±0.950.0003
SHS exposure from smokers can cause lung cancer in non-smokers3.05±1.622.93±1.683.21±1.520.0003
SHS exposure from smoker is harmful to infant and children’s health3.98±1.173.91±1.194.06±1.130.0262
SHS exposure from smoker is harmful to adults’ health3.75±1.233.67±1.273.85±1.150.0093
Mean total score of combined smoking and SHS exposure related knowledge3.89±0.903.82±0.933.97±0.860.0039

* p for Student’s t-tests.

#### Awareness of e-cigarettes

This was assessed by asking a single-item question, ‘Have you ever heard of electronic cigarettes prior to this survey?’ with a brief description of e-cigarettes and a picture showing different types of e-cigarettes (with a ‘yes/no’ forced choice format response).

### Statistical analyses

Descriptive statistical analyses were conducted for sociodemographic variables and scores of perceptions of risk for smoking. Means and standard deviations (SD) were calculated for continuous data, while percentages were calculated for categorical data. Student’s t-test was used to compare the smoking risk perception difference between smokers and caregivers and between the participants who had different risk perceptions about e-cigarettes (e.g. e-cigarettes are less harmful than combustible cigarettes vs e-cigarettes are equally or more harmful than combustible cigarettes). The χ2 and Fisher’s exact probability tests were used to compare the perception of e-cigarettes between different sociodemographic groups. Logistic regression was used to determine predictors for e-cigarette harms, benefits, and addictiveness, adjusting for demographic and other characteristics. All collected data were analyzed with SAS 9.4 (SAS Institute, Inc., NC). The level of p<0.05 was considered statistically significant in all tests.

## RESULTS

Of the households approached, a total of 891/1142 (78%) in Taizhou and 320/356 (90%) in Dali participated in the survey (81% overall participation rate). Of the total (N=1211) household members who participated in the study, 668 were smokers and 543 were non-smokers. Of the participants (N=1211), a higher proportion (74%) was from the Taizhou region, the mean age was 44.2 (±13.54) years (range: 20–81), 58% were male, 34% were farmers and most (70%) had an education level of middle school. Almost all of the smokers were male (99%), while most of the non-smokers were female (97%); most participants were married (98%) (Table 1).

### Smoking and secondhand smoke exposure-related combined knowledge among participants

As shown in Table 2, non-smokers had significantly higher combined smoking and SHS exposure-related knowledge scores (3.97±0.86) (range: 1–5) than smokers (3.82±0.93) (p<0.01). Non-smoker’s knowledge scores were significantly higher in the items of ‘Smoking increases the risk of heart diseases’ (3.85±1.18 vs 3.65±1.34) and all items about SHS exposure-related knowledge. But no significant difference in the perceived risk of COPD and lung cancer were found between non-smokers and smokers (Table 2).

### Awareness and perceptions about e-cigarettes

As shown in Supplementary file Table S1, a higher proportion of smokers (57%) than non-smokers (47%) knew about e-cigarettes (p<0.01). Among those participants who knew about e-cigarettes (n=604), a significantly higher proportion of non-smokers (18%) than smokers (10%) thought e-cigarettes are addictive (p<0.01).

As shown in Table 3, of those who knew about e-cigarettes, a significantly higher proportion of participants from Dali (77%) than Taizhou (59%) and those who were of ethnic minority (i.e. not Chinese Han) (76%) than Chinese Han (59%) thought e-cigarettes were less harmful (p<0.02). Participants who had a higher perceived risk (4.37±0.90) on the item ‘Smoking increases the risk of COPD’ thought e-cigarettes were less harmful than those who had a lower perceived risk (4.18±1.11) (p<0.05) (Table 4). Participants who had higher mean knowledge scores on SHS exposure-related items, including ‘Concerned about the harmful effects of SHS exposure to children’s health’, and ‘SHS exposure from smoker is harmful to adults' health’, also thought e-cigarettes were less harmful (Table 4).

##### Table 3

Association between demographic factors and perceptions of e-cigarette use in rural areas of Taizhou and Dali, China 2018

FactorsRelative harmfulness (n=553)
Benefits (n=598)
Less harm n (%)Equal or more harm n (%)pHelpful to quit smoking n (%)Otherwise n (%)pYes n (%)Otherwise n (%)p
Region0.00860.49060.7817
Taizhou294 (59)206 (41)284 (56)219 (44)68 (13)439 (87)
Dali41 (77)12 (23)50 (53)45 (47)12 (12)85 (88)
Age (years)0.67870.33510.0608
≤34162 (62)99 (38)169 (59)119 (47)35 (12)254 (88)
35–5498 (59)67 (41)92 (52)86 (48)31 (17)150 (83)
≥5562 (57)46 (43)63 (56)50 (44)9 (8)105 (92)
Gender0.90470.18450.0056
Male199 (60)131 (40)198 (54)171 (46)38 (10)337 (90)
Female135 (61)87 (39)135 (59)93 (41)41 (19)187 (81)
Ethnicity0.01930.45170.9858
Chinese Han297 (59)206 (41)287 (57)221 (43)68 (13)445 (87)
Minority38 (76)12 (24)47 (52)43 (48)12 (13)79 (87)
Marital status0.3924*0.8380*0.8168*
Married327 (61)211 (39)324 (56)257 (44)77 (13)510 (87)
Single3 (38)5 (62)6 (60)4 (40)1 (10)9 (90)
Divorced or widowed3 (60)2 (40)2 (40)3 (60)1 (20)4 (80)
Occupation0.83110.75470.9599
Farmer88 (61)56 (39)88 (52)81 (48)24 (14)147 (86)
Self-employed71 (63)41 (37)70 (57)52 (43)16 (13)108 (87)
Unemployed42 (55)34 (45)46 (61)30 (39)11 (14)66 (86)
Worker76 (60)51 (40)74 (57)55 (43)15 (11)118 (89)
Other57 (63)34 (37)55 (57)42 (43)12 (13)82 (87)
Education level0.20580.30240.7182
Primary school or below55 (53)48 (47)56 (54)48 (46)13 (12)92 (88)
Middle school124 (64)70 (36)120 (53)106 (47)27 (12)205 (88)
High school or above151 (61)96 (39)154 (60)104 (40)36 (14)220 (86)

* Fisher’s exact probability test.

##### Table 4

Association between smoking and SHS exposure related knowledge, and perceptions of e-cigarette use, rural areas of Taizhou and Dali, China 2018

VariablesRelative harmfulness
Benefits
Less harm mean±SDEqual or more harm mean±SDpHelpful to quit smoking mean±SDOtherwise mean±SDpYes mean±SDOtherwise mean±SDp
Smoking increases the risk of lung cancer4.39±0.924.29±1.050.27824.49±0.874.14±1.11<0.00014.35±0.964.33±1.010.8605
Smoking increases the risk of COPD4.37±0.904.18±1.110.03314.41±0.904.16±1.060.00274.31±0.954.29±1.000.8766
Smoking increases the risk of heart diseases4.03±1.153.86±1.180.10914.09±1.123.78±1.260.00203.96±1.203.95±1.190.9133
Concerned about the harmful effects of SHS exposure to children’s health4.48±0.914.30±1.080.04034.52±0.894.30±1.070.00744.59±0.954.39±0.980.0967
SHS exposure from smokers can cause lung cancer in non-smokers3.25±1.533.10±1.610.27113.33±1.502.99±1.680.01053.38±1.473.13±1.610.2106
SHS exposure from smoker is harmful to infant and children’s health4.23±0.864.12±1.070.18304.28±0.874.10±1.040.01894.31±0.944.19±0.940.2657
SHS exposure from smoker is harmful to adults’ health4.06±0.873.87±1.140.03024.10±0.913.81±1.130.00103.98±1.013.97±1.010.9621
Mean total score of smoking and SHS exposure related knowledge4.15±0.664.03±0.810.09474.21±0.663.97±0.780.00024.19±0.644.08±0.730.2609

[i] *p for Student’s t-tests.

The mean total combined score of smoking and SHS exposure-related knowledge in participants who thought e-cigarettes are helpful to quit smoking was higher than in those who did not think e-cigarettes can help in quitting (p<0.05). Participants who thought smoking could increase the risk of lung cancer, COPD, or heart diseases, were more likely to believe that e-cigarettes were helpful to quit smoking than those who did not believe so (p<0.05 for all). Participants who had higher mean knowledge scores on SHS exposure-related items, including ‘Concerned about the harmful effects of SHS exposure to children's health’, ‘SHS exposure from smokers can cause lung cancer in non-smokers’, ‘SHS exposure from smoker is harmful to infant and children's health’ and ‘SHS exposure from smoker is harmful to adults' health’, were also inclined to believe that e-cigarettes were helpful to quit smoking (Table 4). No demographic variable was significantly associated with perceptions of e-cigarette benefits (Table 4).

The only difference in the perceptions about e-cigarette addictiveness was found between different genders. A higher proportion of females (19%) than males (10%) perceived that e-cigarettes are addictive (p<0.01) (Table 3). None of the items about smoking and SHS exposure-related knowledge was significantly associated with perceptions of e-cigarette addictiveness (Table 4).

The multivariate logistic regression models including all the significant variables from Tables 3 and 4 (i.e. region, gender, ethnicity, risk of lung cancer, COPD and heart disease from smoking, risk of lung cancer from SHS exposure, and concerned about harmful effects of SHS to infant, children and adults’ health) and adjusting for region, sex, age, education level and occupation, identified only one variable associated with the benefits of e-cigarette use (i.e. e-cigarettes are helpful to quit smoking) that was associated with holding a higher knowledge about smoking and SHS exposure (OR=0.608; 95% CI: 0.450–0.820) (Table 5).

##### Table 5

Multivariate logistic regression analysis of perceptions of e-cigarette use among Chinese in rural areas of Taizhou and Dali (smokers and non-smoker)*

VariablesRelative harmfulness (equal or more harm vs less harm)
Benefits (helpful to quit smoking vs otherwise)
Smoker
Non-smoker
Smoker***
Non-smoker
Smoker***
Non-smoker
OR (95% CI)OR (95% CI)OR (95% CI)OR (95% CI)OR (95% CI)OR (95% CI)
Knowledge about smoking and SHS exposure (total mean score)*0.704 (0.514–0.965)1.204 (0.707–2.051)0.608 (0.450–0.820)0.676 (0.403–1.135)1.045 (0.624–1.751)1.158 (0.589–2.278)
Knowledge about smoking and SHS exposure (total mean score)**0.711 (0.520–0.973)1.289 (0.761–2.184)0.607 (0.450–0.819)0.695 (0.417–1.156)1.080 (0.645–1.808)1.050 (0.546–2.016)

* Models were adjusted for region, sex, age, education level and occupation.

** Models were adjusted for region, age, education level and occupation.

*** Model is statistically significant (p<0.05).

## DISCUSSION

Findings from this study deepened our understanding of rural household adults’ self-reported usefulness and perceptions of e-cigarettes. The findings revealed distinct perceptions of e-cigarettes addictiveness among smokers and non-smoker caregivers as well as among men and women in the rural households with children. We only had three female smokers in our sample, which is obviously lower than the rural female smoking rate (2.2%, 95% CI: 1.5–3.2) as reported in the 2010 census25. Also, our small number of female smokers may be because our subjects were from households with children aged ≤5 years; many mothers might have quit smoking while pregnant or after giving birth.

We found that the awareness of e-cigarette use among our rural population sample is relatively high. Consistent with a previous study that found smokers (71.9%) than non-smokers (46.8%) had a much higher awareness of e-cigarettes26; our results showed that the awareness rates were 59% among smokers and 47% among non-smokers. It should be noted that, in this study, we only collected a positive or negative response of awareness of e-cigarettes instead of obtaining in-depth information about actual usage of e-cigarettes and how the awareness differs based on the status of e-cigarette use. Future studies should evaluate the relationships of e-cigarette awareness and perceptions with the actual usage of e-cigarettes. Also, our subjects are parents or household members with children at home. Having a child at home might have influenced participants’ perceptions and understanding about the hazards of cigarette smoking as well as e-cigarette use in consideration of the health of the child at home and the potential impact on the child to become a future cigarette smoker or e-cigarette user.

In an earlier study, 52.6% of the Chinese reported that e-cigarettes assist to quit smoking26. In the current study, we also found that over half of the participants perceived e-cigarettes as helpful in quitting smoking (56.1%) and less harmful (55.2%) relative to combustible cigarettes. Also, our results correspond to the summary findings from 18 Chinese e-cigarette websites that reported the following most common beneficial claims of e-cigarettes: health-related benefit (89%), no SHS exposure (78%), and helpful to quit smoking (67%)27. It can be expected that commercials of e-cigarettes introduce more positive perceptions of e-cigarettes, which may eventually lead to a higher prevalence of e-cigarette use. This influence might be particularly strong among smokers as they may be more sensitive to tobacco-related advertisements. However, very few e-cigarette products have been fully evaluated for their toxicological properties28, let alone in China where regulations over e-cigarettes are nearly non-existed so far. Although some studies have claimed possible benefits of using e-cigarettes29, they are not harmless and the long-term effects of their use are still unknown30. Considering the unregulated e-cigarette market in China and the lack of studies to assess the long-term impact of e-cigarettes on health, claims of health benefits and their effect on smoking cession should be postponed until these issues have been sufficiently studied. It is also critical to develop health campaigns to educate people about the facts of e-cigarette use, explaining what is known about the use of e-cigarettes as a possible smoking cessation tool and what are the potential health hazards. The World Health Organization also concluded that the evidence is insufficient to recommend e-cigarettes for quitting, and clinicians should direct patients to other cessation aids that are proven to be effective and safe31.

People tend to think e-cigarettes are less addictive than combustible cigarettes32. In this study, non-smokers had significantly higher awareness (38.7%) of the addictiveness of e-cigarettes than smokers (24%), though the general awareness was still very low. Although our knowledge of nicotine dependence from e-cigarettes is still limited33, nicotine in e-cigarettes may introduce addiction34.