This study used data from the “COVID-19 Effects on the Mental and Physical Health of Asian Americans & Pacific Islanders Survey Study” (COMPASS), a community-based national survey that assesses COVID-19 effects on AANHPI adults. Data collection of the first survey (COMPASS I) occurred during October 2020 and January 2021, with 5420 participants completing the survey. Details about participant recruitment and study procedures have been reported elsewhere [20]. Briefly, eligibility criteria included self-identification as AANHPI alone or in combination with other races and ethnicities, being able to read English or Chinese (traditional and simplified Chinese) or Korean or Samoan or Vietnamese (the survey languages), being 18 years old or older, and residing in the United States. Participants were recruited through community partners who served AANHPI communities, personal or professional networks, social media, ethnic media, and through the Collaborative Approach for Asian Americans, Native Hawaiians, and Pacific Islanders Research and Education registry [21]. Participants completed the survey by themselves via the study website or were interviewed by research staff over the phone or in person. Data were collected using REDCap (Research Electronic Data Capture; Vanderbilt University) [22].

COMPASS I participants were invited to complete a follow-up survey (COMPASS II) between December 2021 and May 2022. This study analyzed data from the 3411 participants who completed the COMPASS II survey.

The institutional review board at the University of California San Francisco reviewed and approved the COMPASS study (UCSF IRB: #20‐31925). Participants were provided with detailed information about the study, and informed consent was obtained before survey completion. Participants received a US $10 gift card upon survey completion. Personal details of participants were kept confidential, and a unique study identification number was assigned to each participant for data management. Deidentified data were used for the analysis.

Data were analyzed from April 15 to May 30, 2024, using Stata version 17 (StataCorp LLC). Descriptive statistics were used to describe the characteristics of participants and substance use. Multivariable logistic regression models were used to examine the association between each type of substance use and the cultural group and other covariates for each sex. To compare the odds of using a substance between a cultural subgroup and the rest of the AANHPI adults, we repeated regression models for the dummy variable of each cultural subgroup. All statistical tests were 2-sided, and P<.05 was considered statistically significant.

Participant characteristics are provided in Table 1. Among 3411 participants, the majority self-identified as ethnic Chinese (n=1176, 34.5%), Korean (n=692, 20.3%), and Vietnamese (n=669, 19.6%). Participants from Japanese, Asian Indian, Filipino, and NHPI culture groups each accounted for less than 5% of the sample. Most participants were female (n=2262, 66.3%), had a bachelor’s degree or higher (n=2527, 74.1%), were born abroad (n=2175, 63.8%), and reported recent discriminatory experiences (n=2050, 60.1%).

The proportions of current substance use among AANHPI adults by cultural groups and sex are presented in Table 2. The overall prevalence of cigarette smoking among AANHPI adults was 4.2% (142/3359), with 8.4% (94/1123) of males and 2.1% (48/2236) of females reporting current smoking. Among male AANHPI adults, the prevalence of current cigarette smoking varied widely from 3.7% (2/54) among Asian Indians to 9.8% (21/214) among Vietnamese and 14.8% (10/71) among the multicultural group. The prevalence of current cigarette smoking among AANHPI females was low (<3%) in most cultural subgroups, except for NHPI (4/61, 6.6%) and those from the multicultural group (8/153, 5.2%).

The overall prevalence of cannabis use was 5.5% (184/3235) among AANHPI adults, with the prevalence almost equal between males and females. Cannabis use prevalence was highest among Filipinos (12/113, 10.6%) and multicultural participants (29/221, 13.1%). Among males, cannabis use ranged widely from 1.9% in participants of Asian Indian (1/53) and Vietnamese (4/211) cultures to 15.7% (11/70) in those of the multicultural group. Cannabis use was highest among females of the multicultural group (18/151, 11.9%) and Filipino (9/82, 11%). In addition, cannabis use was more prevalent among females than males for participants of Vietnamese (5.7% [25/442] vs 1.9% [4/211]) and Asian Indian (4% [4/100] vs 1.9% [1/53]) cultures.

Alcohol use was common among AANHPI adults (1265/3361, 37.6%) and more common among males (505/1126, 44.8%) than females (760/2235, 34%). Among cultural subgroups, the prevalence of current alcohol use varied from 7.4% (7/95) among NHPIs to 59.7% (68/114) among Filipinos. Male participants of Asian Indian (58/153, 54.7%), Korean (141/265, 53.2%), or a multicultural group (40/71, 56.3%) were most likely to report current alcohol use. Among females, alcohol use was most common among Filipinos (44/82, 53.7%), Koreans (175/414, 42.3%), Japanese (48/116, 41.4%), and the multicultural group (64/151, 42.4%).

Among male participants, there were no statistically significant differences in the odds of cigarette smoking between each cultural subgroup compared with the rest of the AANHPI adults (Table 3). However, compared with the rest of the AANHPI males, Vietnamese males were less likely (adjusted odds ratio [AOR] 0.26, 95% CI 0.09‐0.77) to report cannabis use, whereas those males of the multicultural group had higher odds of reporting cannabis use (AOR 2.39, 95% CI 1.10‐5.19). In addition, Korean males had higher odds of alcohol use (AOR 1.41, 95% CI 1.02‐1.95) compared with other males, while Ethnic Chinese males were less likely to report alcohol use (AOR 0.63, 95% CI 0.48‐0.83) compared with their other male peers.

The perceived discriminatory experience was significantly associated with higher odds of smoking (AOR 2.08, 95% CI 1.19‐3.61) and drinking alcohol (AOR 1.31, 95% CI 1.01‐1.72) but not using cannabis. Male participants with moderate or severe depression and anxiety symptoms were almost 3 times more likely to report smoking (95% CI 1.42‐5.51) and 5 times more likely to report cannabis use (95% CI 2.58‐11.37) compared with those who had a normal level of depression and anxiety symptoms. Being born in the United States was associated with higher odds of drinking alcohol, but not for cigarette and cannabis use. English proficiency was not found to be associated with using any of these 3 substances. In addition, male participants of older age groups were less likely to report cannabis use than young adults aged 18‐29 years, and those with higher levels of education were also less likely to report smoking and cannabis use compared to those with less than high school education.

Factors associated with substance use among female participants are shown in Table 4. The odds of smoking among female Koreans were twice (AOR 2.27, 95% CI 1.04‐4.98) that of other non-Korean AANHPI females. Female participants of the multicultural group were about 2 times more likely to report cannabis use (AOR 1.89, 95% CI 1.06‐3.37) compared with the rest. Female participants with Filipino (AOR 1.84, 95% CI 1.15‐2.93) and Korean (AOR 1.70, 95% CI 1.32‐2.19) cultural backgrounds were more likely to report alcohol use, while those females with ethnic Chinese, Vietnamese, and NHPI backgrounds were less likely to report alcohol use compared with females of the rest.

The perceived discriminatory experience was significantly associated with smoking (AOR 2.21, 95% CI 1.02‐4.80) among female participants but not for other substances, whereas depression and anxiety symptoms were not associated with smoking. However, female participants with elevated depression and anxiety symptoms were more likely to report using cannabis and alcohol than those females with a normal level of depression and anxiety symptoms. Being born in the United States was significantly associated with a higher likelihood of using all 3 substances among AANHPI females. Older females were more likely to smoke cigarettes than young females between ages 18 and 29 years, but the relationship was the opposite for cannabis and alcohol use. Limited English proficiency was significantly associated with lower odds of alcohol drinking (AOR 0.53, 95% CI 0.37‐0.75) but not cigarette and cannabis use. In addition, females of older age groups were more likely to report smoking but less likely to report cannabis and alcohol use compared to their young adult counterparts aged 18‐29 years. Female participants with higher levels of education were also less likely to report smoking and cannabis use compared to those with less than high school education.

Given that disaggregated data about substance use by sex among the AANHPI communities are scarce, this study adds several important findings to the literature. First, our study showed different patterns of use of 3 substances, including cigarettes, cannabis, and alcohol, among a nationally recruited sample of AANHPI adults. Second, the study provided disaggregated data for each type of substance use and indicated a wide variation in the prevalence of use by cultural subgroups within and between each sex. For example, smoking was mostly among AANHPI males and ranged from 3.7% among Asian Indians to 14.8% among the multicultural group; whereas, the rates of cannabis use among AANHPI adults were similar between the 2 sexes, and Filipinos reported the highest rate of current cannabis use. Alcohol use was common among males of some subgroups, including Filipino, Indian, and Korean. Third, findings revealed different factors associated with each type of substance use within each sex.

Despite a low overall prevalence of cigarette smoking (4.2%) among AANHPI adults, the prevalence among AANHPI males was almost four times higher than among AANHPI females. The sex disparity in smoking remained consistent with findings from previous studies [16,17]. In addition, smoking prevalence varied widely across cultural subgroups among male AANHPI adults. Particularly, participants of Vietnamese (9.8%) and multicultural subgroups (14.8%) had the highest prevalence of current smoking, which was comparable to estimates for general US adults in 2022 (11.6%) or adults of other racial/ethnic groups such as non-Hispanic White (12.7%), non-Hispanic Black (14.2%), and Hispanic (8.0%) [25] of the National Health Interview Survey. Our finding confirms the need for interventions focusing on smoking among high-risk subgroups of AANHPI males, given that cancer is the leading cause of death and lung cancer is the leading cause of cancer mortality among Asian American men [26].

We found that experience of discrimination was significantly associated with current smoking among both male and female AANHPI adults, which is consistent with previous studies conducted before and during the COVID-19 pandemic [27-29]. Some AANHPI adults might have adopted smoking as it may have been perceived as a way to cope with increased discrimination experience that they faced during the pandemic [30]. In addition, racial discrimination may also contribute to depressive symptoms by increasing internalized negative self-perceptions and lowering self-esteem [31]. However, elevated levels of depression and anxiety symptoms were significantly associated with higher odds of smoking among AANHPI males only but not among AANHPI females in our study. Current literature shows mixed results in the association between sex, depressive symptoms, and smoking, with some studies reporting a significant smoking-depression association in women but not in men [32-34], some studies indicating a significant association in both sexes [35], and others showing a significant association among males but not in females [36]. These mixed findings could be attributed to the differences in study samples and analysis modeling. Future studies should explore the mechanism for sex differences in the association between smoking and depression.

Regarding cannabis use, we found that Filipino and multicultural groups had the highest rate of cannabis use. The finding of a high rate of cannabis use among Filipinos was consistent with the findings from another study which also found Filipinos had the highest rate of past 12-month drug use (eg, cannabis, cocaine, prescription drugs without physician recommendation, and other drugs) among Asian Americans [19]. The overall prevalence of cannabis use among NHPIs in our study was 5.3%. There are limited studies on cannabis use among NHPIs [37], including one study conducted during the COVID-19 pandemic (between April and November 2021) reporting the rate of lifetime use of illicit substance use (eg, cannabis, opioids) of 35% among NHPIs [18]. Given the small sample of the NHPI group and the underrepresentation of NHPI males in our study, future studies should re-estimate the prevalence of cannabis among NHPIs.

Previous studies that do not take into account ethnic subgroups often reported lower rates of drug use (including cannabis) among female Asian Americans than their male counterparts and females of other race or ethnic groups [19,37]. However, female participants of Asian Indian and Vietnamese groups even seemed to have a higher rate of cannabis use than their male counterparts in our study, indicating the need for taking into account both ethnic subgroups and sex differences in cannabis use among AANHPI adults. We found that younger age (ie, 18‐29 y) and elevated levels of depression and anxiety symptoms were associated with a higher likelihood of cannabis use in both male and female AANHPI adults, which was also consistent with the previous study [19]. However, in our study, discrimination experience and English proficiency were not associated with cannabis among AANHPI males or females, and being born in the United States was only associated with cannabis use among female AANHPI adults.

Similar to smoking and cannabis use, national data (ie, National Survey on Drug Use and Health) often indicate that Asian Americans have the lowest estimated current alcohol use compared to other racial groups (eg, 40.3% among Asian Americans vs 57% among Whites, 45.9% among African Americans, and 49.2% among Hispanics) [38]. Although the overall prevalence of alcohol use among AANHPI adults in this study (37.7%) was comparable to the rate reported from the national data, our study showed that some subgroups (eg, Filipinos or Koreans) had the current alcohol use rate as high as other racial groups. Previous studies also indicate important correlates of alcohol drinking are racial discrimination experience, depression and anxiety symptoms, and nativity [30,39]. This study added more insights to the variation of these correlates by sex, such as a significant association between alcohol use and discrimination experience among AANHPI males but not among AANHPI females.

This study has several limitations. First, despite a large nationally recruited sample of AANHPI adults, the generalizability of the findings to all AANHPI adults across the United States was limited due to a nonprobability sampling approach. In addition, we were unable to estimate the prevalence of substance use among males of some cultural subgroups (ie, Japanese, Filipino, and NHPI) due to a small sample of these groups. Data from this study were collected during the COVID-19 pandemic, so the estimated prevalence of substance use for participants could be lower than when measured at a normal time. Finally, self-reported data are subject to recall and social desirability bias.

Our study reconfirmed the need to address the diversity of AANHPI adults in public health programs, including programs that address substance use and health-related disparities to address various needs across cultural subgroups and general AANHPI adults who experience mental health issues and discrimination. A high rate of substance use among some subgroups of the AANHPI adults has contributed to the disparity in their health outcomes, including lung and liver cancers. Also, differences in cultural norms, stigma, and language barriers could prevent AANHPI adults from accessing social support and treatment services [40]. Yet, AANHPI populations remained underrepresented in clinical research, including the development of substance use prevention and treatment programs [12,41]. Thus, it is essential to develop and implement culturally appropriate prevention and cessation programs to address the needs of AANHPI adults. These programs should engage communities in their development to ensure that intervention content, languages, and delivery methods are aligned with the cultural beliefs, values, norms, and preferences of specific AANHPI groups [41].