University students from 63 universities and colleges in the Jilin province, China, were invited to participate in a cross-sectional web-based self-report survey. Recruitment was performed via universities’ and colleges’ staff teachers and targeted WeChat groups (similar to WhatsApp groups). Students who were interested in participation entered the secure and anonymous questionnaire through the QR code mentioned in the study poster that was distributed by school staff teachers. Students were first asked to thoroughly read the study description, and those who selected the commensurate button and continued to answer the questionnaire were deemed as consented to this study. Individuals who successfully completed this questionnaire were offered the opportunity to join a cash prize lottery by accessing a separate link. The description of the cash prize lottery was listed on the front page of the web-based survey, and the amount was carefully calibrated to strike a balance between being sufficient to encourage participation and minimize dropout without being excessively large to create an undue influence on participants’ behavior or reporting. Respondents’ data were collected between October 26 and November 18, 2021, via Credamo [23].

For statistical quality control of this study, we excluded respondents who (1) were younger than 15 years (ie, below the lowest threshold of the regular age range of college or university students in China), (2) failed 2 or more of the 4 attention check questions (ie, this question is an attention check question. There are many colors of the sky, which can be green, blue, red and black, but, for this question, please choose “green” for your answer) that were designed to assess whether survey respondents were focused during their responding process, and (3) showed presence of logical contradictions or inconsistencies within responses (eg, male participants choosing “male” for sexual preference while choosing “heterosexual” for sexual orientation), omission of answers (eg, leaving blanks in their answers instead of choosing options such as “none of the above” or “prefer not to say”), or their answers showed evidence of patterned or nondiscriminatory responses (eg, consistently choosing the first option across multiple questions). The final sample consisted of 89,342 participants, among which 8853 were SGM respondents and 80,489 were cis-heterosexual respondents.

First, we performed a descriptive analysis of the sociodemographic variables of the SGM and the cis-heterosexual groups. The Mann-Whitney U test and chi-square test were performed for nonnormally distributed continuous variables (eg, scores of CC dependence, EC dependence, NSSI, suicidality) and categorical variables (eg, classifications of cigarette usage, NSSI, and suicidality), respectively. Our primary analyses using the multivariable linear regression model involved examinations of both the influence of CC usage and EC usage on NSSI and suicidality and the interaction effects of CC usage and EC usage on NSSI and suicidality by group (ie, SGM and cis-heterosexual). Both models were controlled for other sociodemographic variables. Results were reported at 95% CIs and were only considered significant when P values were less than .05. For variables with a small number of missing values (eg, age, education background), we used mean substitution to input the missing data. We used ω as the measure of the tools’ reliability index [33]. All statistical analyses were performed using SPSS (version 28, IBM Corp).

This cross-sectional survey received approval from the ethics committee at Jilin University (approval 2021-9-29). Informed consent for primary data collection and secondary data analyses from participants was automatically collected from students who selected the commensurate button and continued to answer the questionnaire. The privacy and confidentiality of the participants in this study were protected during data collection and analysis through anonymization and deidentification. None of the personally identifiable information was collected, and all data were stored in a personal computer in the laboratory with password protection. As mentioned earlier, compensation for participants in the study was provided with a link on the postsurvey completion page to the cash reward lottery, with the amount being meticulously calculated to be rewarding while not generating potential bias.

To our knowledge, this is the first study examining the usage of different cigarette types (ie, CCs and ECs) and their relationship with NSSI and suicidality while investigating intergroup differences between Chinese cis-heterosexual and SGM youth by using data from a large-scale cross-sectional survey. We found that CC usage and dependence were more prevalent among cis-heterosexual individuals, whereas EC usage and dependence were more prevalent among SGM individuals. We found that only CC usage or EC usage increased individuals’ risks for both NSSI and suicidality; in particular, dual usage increased these risks. We also found that CC dependence and EC dependence could simultaneously impact individuals’ risks for NSSI and suicidality independently, and the impacts of each could overlay with the usage of CCs and ECs.

Our findings shed light on the existing intergroup differences in susceptibility to NSSI and suicidality risks. Although CC usage, EC usage, and dual usage increased an individual’s risk of NSSI and suicidality compared with no usage, we found that only CC usage and dual usage increased the risks of NSSI and suicidality in the SGM population compared to those in the cis-heterosexual population. These results are supported by previous research [34-42]. SGM youth are susceptible to the disproportionately high risk of CC usage due to societal, peer, and community influence, which are the major risk factors for NSSI and suicidality. Such societal, peer, and community influences could also be applied to EC usage with its increasing advertisements on social media, and EC usage could be a significant risk factor for mental health risks. Furthermore, recent studies [34-43] underline the fact that there is a significant overlap between CC and EC users in the SGM population, supporting our finding that individuals with dual usage have the highest risks of NSSI and suicidality compared to individuals with no usage of CC and EC.

We found that both CC dependence and EC dependence increased an individual’s risk of NSSI and suicidality. Specifically, we found that both CC and EC dependence increased the risks of NSSI in the SGM population compared to that in the cis-heterosexual population. However, only EC dependence increased the risk of suicidality in the SGM population compared to that in the cis-heterosexual population. Although only a relatively limited amount of research has focused on investigating EC dependence, findings showed that youth with EC usage could experience dependence symptoms (ie, higher usage frequency compared to CC) unique to ECs, creating more barriers to EC cessation [43]. Considering that SGM youth could be more influenced by EC due to societal, peer, and community influence and EC has been deemed as the major risk factor for SGM population’s mental health, we believe our findings on EC dependence and its impact on SGM youth could be valuable for further research. Thus, our findings add to the emerging breadth of evidence emphasizing the imminent risks posed by CC and EC usage and dependence among youth.

The existing intergroup difference between SGM and cis-heterosexual college students in NSSI and suicidality risks resulting from CC and EC usage and dependence can be explained by SGM individuals’ less stable mental health and their vulnerability toward CC usage and EC dependence. First, findings from US and international studies have consistently underlined the continuing worsening situation of SGM youth’s mental health [34-37]. Such a disparity arises from a wide range of structural vulnerabilities such as unavoidable systemic discrimination and prejudice in education, employment, and health care opportunities [38,39]. Further, these structural vulnerabilities stem from the imbalanced hierarchical social power structure that is responsible for generating and widening the mental health gap between SGM and cis-heterosexual individuals while limiting SGM individuals’ accessibility to determinants of health, leaving the SGM population defenseless against risky behaviors such as substance use [39-45].

As previously mentioned, a number of explanations have been proposed in past research, including being impacted by the marketing target of the tobacco industry on the media scale, community norms on a group scale, and minority stressors on a personal scale. First, it is well-established that SGM youth are at disproportionately high risk of CC usage [41,42]. Past research also suggests that the SGM population has been the major marketing target of the tobacco industry, resulting in increased difficulties for SGM youth to resist CC usage when such behavior is normalized by the public media [44]. With the increase in EC usage, recent research underlines that the same aggressive marketing strategy has also been employed by the EC industry and public media. To be more precise, studies indicate that SGM youth were more likely to report being exposed to, searching for, sharing, or being shared with EC–related content on the news or social media platforms compared to their non-SGM peers [46-50]. Second, from a community standpoint, past research has pointed out that the community context could be an influential factor in SGM youth’s smoking behavior [47-50]. Under the influence of increasing exposure to EC-related content among the SGM population, it becomes reasonable to assume that, in social situations with SGM peers (ie, one of the potential triggers for EC usage and combined CC and EC usage), SGM youth would become more susceptible to the rising trend of EC usage among their peers [48]. Third, on a personal scale, as mentioned earlier, the minority stress model underscores that unique stressors experienced by the SGM population, including but not limited to discrimination and internalized homo/bi/transphobia, could result in higher stress levels among this stigmatized group [40-45]. Such stress could then contribute to the thought pattern of EC usage being a way for individuals to express or assert their gender identity or to rebel against traditional gender norms [40-45]. Due to all the aforementioned reasons, concerted efforts need to be made on a societal level in terms of educating SGM youth about the misconceptions, safety concerns, and, most importantly, adverse mental health outcomes such as increased risks for NSSI and suicidality that could be potentially induced by EC usage.

From a broader perspective, the rise of EC usage among Chinese youth has become a public health concern in recent years. Unlike the easy accessibility of the adverse effects of CC usage from various reliable sources, youth often receive inaccurate information about EC usage, resulting in a higher likelihood of maintaining their positive perception of ECs. Under the massive impact created by extensive marketing campaigns, ECs have been promoted as less harmful than CCs and, therefore, have rapidly gained popularity among the younger generation [45]. Compared to the downhill journey of CC usage among Chinese youth (ie, a significant reduction from 12.9% in 2014 to 3.9% in 2019), the rise in EC usage (ie, from 1.2% in 2014 to 2.7% in 2019) reveals that smoking among youth is still a major public health concern in China, many of whom are not fully aware of the physical and mental health risks induced by EC usage [46]. Although camouflaged by the utility for smoking cessation, nonsecondhand smoke exposure, and easy accessibility, ECs’ negative impact on youth’s physical and psychological well-being cannot be ignored. Considering EC’s relationship with NSSI and suicidality, efforts to prevent youth from EC addiction must be made on a societal level [50].

It is worth noting that Chinese government officials and policy makers have promulgated the Law on Youth Protection and published the Notice of the “Guarding Youth” Special Action Plan for Youth Protection from Tobacco Abuse in 2021 while further strengthening the supervision of EC products, regulating the market order, and standardizing industrial governance by putting into effect the Measures for the Administration of Electronic Cigarettes as of May 1, 2022 [50]. The implementation of these measures has prohibited youth younger than 18 years from purchasing EC products, which could correspondingly reduce their NSSI and suicidality risks. Nevertheless, to prevent the situation from further worsening, future efforts are still required from a societal level to regulate the aggressive marketing of the EC industry and to expand the media and school coverage of educational campaigns in formats such as nonprofit advertisements, school lectures, and short videos on social media platforms on EC usage prevention and its adverse mental health effects on youth.

The findings from this study must be considered in light of a few limitations. First, findings from this study are limited to Chinese SGM college students and, thus, may not be generalizable to SGM youth of other cultural backgrounds or who reside in other countries. Nevertheless, our findings provide crucial insights into the disparity between SGM and cis-heterosexual youth with regard to EC usage, propelling future studies to further investigate the extent and nature of such disparity. Second, although the effect size found in this study was relatively small, our large sample size indicates that such an effect was still stable on a large population scale. Specifically, the impact of cigarette usage and dependence on suicidality and NSSI in SGM youth is more pronounced than that on their cis-heterosexual peers, highlighting the need from a public health standpoint to pay more attention to the impact of EC and CC on suicidality and NSSI among the SGM population. Third, due to the cross-sectional nature of this study, only associations were examined. Longitudinal data are required to investigate these associations over time. Fourth, the sex assigned at birth of the SGM participants in this study was mostly female (6531/8853, 73.8%), which could potentially incur sampling bias and limit the study findings’ generalizability.

Overall, our findings have crucial implications for understanding the disparity of NSSI and suicidality risks between SGM and cis-heterosexual youth. First, this study contributes to the growing body of literature highlighting the mental health risks induced by CC usage and EC usage, emphasizing that SGM youth are more susceptible to EC usage. Second, since the sample in this study consisted of Chinese college students, these results could suggest that relevant school policies (eg, CC-free and EC-free campus policies that prohibit EC usage in school public areas and dormitories) and other school-level activities (eg, informative workshops directed by student health centers on EC usage, SGM susceptibility to EC usage, and its physical and mental health risks) could be enacted and hosted to address the health disparity among SGM students. Feedback surveys could be distributed for those policies and school-level interventions to ensure that these interventions attenuate SGM students’ needs rather than perpetuate and exacerbate SGM stereotypical views. Third, findings from this study could inform governmental officials and policy makers on the vulnerability of SGM youth to EC usage. For instance, policy makers could further develop comprehensive EC control policies that are similar to the Notice of the “Guarding Youth” Special Action Plan for Youth Protection from the Tobacco Abuse in 2021 but with an emphasis on reducing EC usage and content exposure among SGM youth. Overall, our study provides clear guidance for influential parties to develop effective strategies to resolve the rising imminent public health concern of EC usage among SGM youth.