Introduction

Mental illness poses a tremendous health burden, with up to 46% of Americans meeting diagnostic criteria for mental illness in their lifetime [1]. Racial/ethnic disparities in mental health services utilization have been extensively documented [2,3,4,5,6]. Racially/ethnically minoritized groups are less likely to use outpatient mental healthcare [2], specialty mental health providers [7], and receive quality mental health treatment, relative to a comparable population of White Americans [3]. Among those with mental illness, racially/ethnically minoritized groups access care at half the rate of non-Hispanic Whites [8].

Despite having similar levels of mental illness burden [9], Asian Americans are less likely than White Americans to receive quality mental healthcare, which adversely impacts treatment use such as receiving mental health screenings [7, 10], referrals to specialty mental health treatment [5], being diagnosed with a mental health disorder [11], and receiving psychotropic medication [11]. Variability may in part be attributed to stigma and shame [12,13,14,15,16], differential conceptions of mental illness [12, 17, 18], insurance status [19], lack of culturally and linguistically appropriate care [19, 20], and fragmentation of the mental healthcare system [20]; though, these factors may have a varied effect depending on the Asian subgroup. When Asian Americans perceive a need for mental healthcare or have a diagnosis of mental illness, they are less likely than White Americans to receive mental healthcare [5, 10, 21, 22], even when experiencing suicidality [23], a leading cause of death for Asian Americans [24].

Asian Americans are underrepresented in clinical research [25], and there is limited research examining mental health services use among Asian subgroups. Prior data from largely pre-ACA populations has shown that Asian Americans in aggregate had lower rates of mental healthcare utilization compared to White Americans [2, 4, 7]. What research that does exist examining Asian subgroups shows Chinese, Filipino, Vietnamese, and Other Asian Americans have low rates of mental healthcare utilization relative to White Americans [26]. However, due to the age of the data examined, these findings may not be representative of changes to the healthcare landscape [27, 28].

The objective of this study is to provide a much-needed update to national trends in mental healthcare use (2013–2019) among Asian American subgroups, and to examine differences between non-Hispanic White Americans and Asian Americans using the Medical Expenditure Panel Survey (MEPS). This study builds on prior research highlighting the variation that exists among subgroups, which is usually masked by omission of the Asian group due to lack of sample size or aggregation of Asian Americans [29,30,31,32].

Methods

Data Source

This study analyzed cross-sectional data from the 2013–2019 Household Component of the Medical Expenditure Panel Survey (MEPS), which captures demographics, medical conditions, and health service use of non-institutionalized US civilians. MEPS is widely used to provide national estimates of healthcare measures to inform policy and practice [33]. We combined multiple cross-sections (2013–2019) to increase the precision of point estimates. MEPS verifies health service use by cross-matching reported utilization with data from a random sample of survey participants’ medical providers. Details on the MEPS verification methods are described elsewhere [33]. MEPS oversamples policy-relevant populations, such as Asian Americans, to produce nationally-representative estimates of healthcare utilization [34]. This study used publicly available de-identified data and received a determination of not human subjects research by the Boston University Medical Center Institutional Review Board.

Analytic Sample

The analytic sample consisted of adults (≥ 18 years), who self-reported being non-Hispanic White (n = 112,590 hereafter White) and non-Hispanic Asian (n = 10,210, hereafter Asian (overall)). Adults reporting Asian race also self-reported their subgroup, defined by the country of the respondents’ descendants: Asian Indian (n = 2492), Chinese (n = 2044), and Filipino (n = 1685). Due to limited sample size, MEPS categorizes all Asian subgroups not previously mentioned as Other Asian/Native Hawaiian/Pacific Islander (n = 3989, hereafter Other Asian).

Outcome Measures

We examined differences between White and Asian (overall and subgroups) individuals in mental healthcare use in the past year. Treatment was classified into four categories: any mental healthcare use (outpatient visit or psychotropic medication fill), any outpatient mental health treatment (visit to primary care provider or specialty mental health provider), any specialty mental health treatment (visit to psychiatrist, psychologist, counselor, or social worker), and any psychotropic medication fill. Mental health visits included visits for a disorder covered by the International Classification of Diseases, Ninth Edition (ICD-9) codes 291, 292, or 295-314, and Tenth Edition (ICD-10) codes F01-99. Specialty mental health treatment also included visits classified as “psychotherapy or mental health counseling” by the respondent. Similar outcome operationalization has been utilized in past studies [2, 3]. Psychotropic medications were identified using the Multum Medication Lexicon drug classification system [35].

Independent Variables

The primary predictors of interest were Asian race and Asian subgroup (Asian Indian, Chinese, Filipino, or Other Asian). In secondary analyses, to determine whether disparities differed by mental health status, the predictor of interest was an interaction between Asian race/subgroup and an indicator for elevated risk of mental illness. We defined elevated risk of mental illness as heightened depressive symptoms (PHQ-2 scores ≥ 3) or serious psychological distress (K-6 scores ≥ 13). The PHQ-2 and K-6 are both validated measures for identifying individuals with mental illness among ethnically diverse outpatient populations, with the PHQ-2 demonstrating strong sensitivity (87%) and specificity (78%) for detecting major depressive disorder and the K-6 demonstrating strong sensitivity (90%) and specificity (89%) for severe mental illness [36,37,38]. Regression models adjusted for the following covariates: year (2013, 2014, 2015, 2016, 2017. 2018, 2019), age (18–24, 25–44, 45–64, 65+), sex (male, female), time in the USA among foreign born (less than 1 year, 1–5 years, 5–10 years, 10–15 years, 15 years or more, not applicable), region of residence (Northeast, Midwest, South, West), marital status (yes, no), employment status (yes, no), federal poverty level (FPL; < 100% FPL, 100–124% FPL, 125–199% FPL, 200–399% FPL, > 400% FPL), insurance (private, public, uninsured), and education (less than high school graduate, high school graduate, any college, or college graduate). We also included measures of physical and mental health status, which consisted of self-rated mental health and physical health (excellent, very good, good, fair, poor), Patient Health Questionnaire-2 (PHQ-2) scores (0–6), Kessler 6 (K-6) Psychological Distress Scale scores (0–24), SF-12 physical and mental health scores (0–100), presence of work limitation (yes, no), and number of chronic physical health conditions (0, 1, 2+).

Statistical Analysis

First, we compared demographic, clinical, and service use characteristics between White, Asian, and Asian subgroups using t tests and chi-square tests for continuous and categorical variables, respectively. Next, we plotted unadjusted rates of any mental healthcare use from 2013 to 2019 for Whites and Aggregate Asians among those with elevated and low (i.e., not elevated) risk of mental illness.

We then specified multivariable logistic regression models to estimate mental health service use conditional on the primary predictors and covariates, with an interaction between Asian subgroup and an indicator for elevated risk of mental illness, which allows for the prediction of comparisons of rates of mental healthcare use between White and Asian adults with and without elevated risk of mental illness. For interpretability and to overcome potential bias in estimating interaction terms in nonlinear models, we estimated and reported predicted probabilities for each category using the predictive margins methods [39]. We estimated variances for all analyses which allows us to assess differences by group across risk categories, accounting for the complex study design, nonresponse rates of the MEPS, and standardized stratum and primary sampling unit variables across pooled years [40]. Analyses were completed using Stata version 16 (StataCorp, College Station, Texas) and following STROBE reporting guidelines [41].

Results

In unadjusted analyses of outcome variables, Asian individuals (both overall and Asian subgroups) had significantly lower rates of mental healthcare use than White individuals in all four measures examined (Table 1). Asian Indian individuals reported the lowest rates of any mental healthcare use (7.3%), followed by Filipino, Chinese, Other Asian and White groups (9.0%, 9.3%, 10.2%, and 25.6%, respectively).

Table 1 Weighted population characteristics for White, Asian (total), and Asian subgroup adults (≥ 18 years), 2013–2019 Medical Expenditure Panel Survey
Full size table

There were significant differences between Asian (overall and subgroups) and White individuals in variables related to clinical need, demographics, and socioeconomic status (Table 1). Asian adults were generally younger, higher educated, higher income, more likely to be married, privately insured, and to live in the Northeast and West compared to White adults. Regarding health status, Asian groups were significantly less likely to have had any work limitation and less likely to report poor self-rated physical and mental health and multiple chronic physical health conditions (Table 1).

When stratifying by risk of mental illness, Asian individuals with elevated risk (28% in 2013 to 34% in 2019) were approximately half as likely to access treatment compared to White individuals with elevated risk across all years (59% in 2013 to 66% in 2019) (Fig. 1). Asian individuals without lower risk of mental illness also had significantly lower rates of any mental healthcare use across all years (9% in 2013 to 12% in 2019) compared to White individuals (24% in 2013 and 28% in 2019) (Fig. 1).

Fig. 1
figure 1

Unadjusted rates of any mental healthcare use for Whites and Asians with low and elevated risk for mental illness, 2013–2019. Aggregate of mental healthcare use (“any mental healthcare use”), including any mental healthcare use, any outpatient mental healthcare use, any specialty mental healthcare use, and any psychotropic medication fill

Full size image

After adjustment for variables associated with need for treatment, demographics and socioeconomic status, the overall Asian group was significantly less likely to report any mental healthcare use, any outpatient, specialty and psychotropic medication treatment, compared to their White counterparts. These disparities were consistent across Asian subgroups and by level of risk of mental illness (Table 2). Some disparities (e.g., between Filipino individuals with elevated risk and their White counterparts) were similar in magnitude but not statistically significant because of wide confidence intervals (Table 2).

Table 2 Predicted rates of mental healthcare use by White and Asian Americans
Full size table

When focusing on graphic illustrations of the rates of mental health treatment among those with elevated risk of mental illness, the consistency of mental health treatment disparities is apparent across treatment categories among Asian (overall), Asian Indian, Chinese, and Other Asian individuals (Fig. 2).

Fig. 2
figure 2

Predicted probability of mental health use among White and Asian patients at elevated risk for mental illness. Comparative rates of mental health services use between Whites, aggregate Asians, and Asian subgroups showing the percentage of each subgroup utilizing a given service. Reference population: White. Regression models adjusted for age, sex, time in USA, region of residence, employment status, marital status, federal poverty level, insurance, and education. We also included measures of MEPS participant physical and mental health, which consisted of self-rated mental health and physical health, PHQ-2 scores, K-6 scores, SF-12 physical and mental health scores, presence of work limitation, and presence of chronic condition. Bars represent 95% confidence intervals (± 1.96*SE). *p < 0.05

Full size image

Discussion

We analyzed nationally representative data from the MEPS to estimate disparities in mental healthcare service use among Asian Americans. Asian Americans, overall and subgroups, consistently had lower rates of mental healthcare use than White Americans and these disparities persisted after adjustment for socioeconomic, demographic, and health-related variables. Building on prior research showing disparities in mental healthcare use among Asian populations [5, 7, 10, 11, 21, 22, 42], we find that even among those with elevated risk for mental illness, Asian individuals had lower rates of use relative to White individuals.

Enactment of policies that changed insurance coverage and access to mental healthcare, such as the Affordable Care Act (ACA) and the Mental Health Parity and Addiction Equity Act, did not reduce disparities between Asian and White groups. This is consistent with results from the California Health Interview Survey which showed that the ACA was associated with limited to no improvements in healthcare insurance rates among Asian subgroups and no reductions in disparities for access to and use of general healthcare services [43].

Following national trends, many of the Asian subgroups in our sample were younger [44], had higher levels of education [45], insurance [46], and self-reported mental health [47], which are often protective factors against mental illness, and positively associated with mental health treatment. Even after adjustment for these clinical need and sociodemographic variables, there exist disparities between Asian subgroups and White counterparts.

Our findings may be attributed to stigma surrounding mental illness, a well-established barrier to care among Asians (overall) [12], as well as Asian Indians [13], Chinese [14], Filipinos [15], and Other Asian-Americans. Psychological distress and mental illness may be viewed by some members of Asian communities as an indication of personal failings such as a lack of willpower, poor self-control, or inherent weakness [19, 48]. This stigma can also extend beyond the individual, and may be viewed by others as a failure of the family [19]. As a result, some patients attempt to hide or repress their symptoms, struggles, and self-harm [49], resulting in delays in mental health treatment [50].

Acculturation may also affect mental healthcare access. Asian immigrants have lower rates of utilization than US-born Asians, and endorse a greater number of cultural and structural barriers to care [20]. Additionally, Asian immigrants may be more likely to seek alternative sources of support for mental illness and psychological distress such as other community members, alternative medical providers, and spiritual leaders [12, 15, 51, 52], which may contribute to lower rates of mental healthcare usage in this group. In contrast, utilization and perceived helpfulness of mental healthcare increased with every subsequent generation residing in the USA [26]. While we adjusted for time in the USA in our models, this may fail to capture the full extent of acculturative stressors and protective factors, so further exploration is needed.

Clinicians should be mindful of the diversity of the Asian population and be aware of the various barriers that may contribute to these continued disparities in care. Despite misconceptions to the contrary (e.g., model minority myth) [53], Asians still have substantial unmet need for mental healthcare. Indeed, our adjusted findings show low rates of mental healthcare utilization in all Asian populations relative to White individuals despite high levels of educational and socioeconomic achievement, and Asian individuals overall and in subgroups with elevated risk of mental illness have lower rates of mental healthcare use.

Clinicians should also explore symptoms and complaints which differ from traditional Western illness presentations. Asian Americans may have differing risk factors for suicidality [54]. Portions of the Asian community have cultural views which may contribute to lower mental health utilization, including traditional cultural beliefs [12] and low acceptance of mental healthcare, particularly psychotherapy [13]. There may also be attribution of mental illness symptoms to culturally distinct syndromes or symptoms, such as somatization [17, 18]. This may contribute to lower rates of mental healthcare use among Asians, but does not fully explain treatment disparities [55].

At the same time, clinicians must be careful to avoid overgeneralization [56] (e.g., attributing beliefs held by a subset to the population as a whole) and stereotyping [53] (e.g., model minority myth) which can lead to incorrect assumptions about issues faced by individual patients [57] and erosion of the patient-provider relationship. Previous research suggests that many Asian patients delay seeking care until their symptoms become severe [58], which may contribute to avoidable hospitalizations and emergency room use. Among those who can access mental healthcare, Asian patients report more reasons for treatment non-adherence than White patients. If patients choose to avoid future engagement with the healthcare system due to poor care or negative prior experiences (e.g., racism) [59], their choice represents a failure of the system to provide quality care to diverse populations.

Logistical and systemic barriers to mental health care for Asian subgroups, including treatment costs, time away from work, and health system literacy, contribute to treatment disparities [19, 20]. Additionally, steps are needed to address the shortage of clinicians capable of providing linguistically appropriate and culturally humble care [48, 60], which is an important barrier for Asians with low English proficiency (LEP) [61], and Asian patients who are US-born or immigrated to the USA at a young age.

When designing interventions, stakeholders should recognize relevant differences in cultural, socioeconomic, and linguistic needs in the Asian population and model their efforts on culturally appropriate measures to address these issues [57, 62,63,64]. We highlight other researchers’ caution against aggregating dissimilar populations [29,30,31], and recommend disaggregation of Asian subgroups whenever possible. Given the heterogeneity of the Asian American community, interventions should be made accessible to those of different backgrounds, such as through providing financial resources to those of lower socioeconomic backgrounds, or services for those who may not speak English. Clinicians working with these populations should strive to understand the unique forms of trauma and resilience that are associated with different ethnic subgroups’ histories of immigration and acculturation to the USA. Though there has been increased attention on Asian American mental health in light of increased racist rhetoric and hate crimes in addition to the stress of COVID-19, Asian Americans have long struggled to access timely, quality, culturally sensitive care [65].

Limitations

This research has several limitations. First, small sample sizes limit the precision of our estimates. Second, while the MEPS is a nationally representative survey that captures mental health service utilization and allows for stratification by Asian subgroups, it aggregates many ethnicities into one category: “Other Asian.” This category aggregates heterogeneous and relatively well-studied populations (e.g., Korean Americans, Vietnamese Americans, Japanese Americans) with less researched Pacific Islander and Native Hawaiian populations, each with distinct cultures and barriers to care. The study also neglects important information about cultural, linguistic, and ethnic subgroups by categorizing individuals based on the country of their descendants. Given the lack of data on subgroups within the Asian race and because of the political importance of identifying groups in need of treatment, we feel our analyses remains important, but nonetheless recognize the heterogeneity across the Asian diaspora that informs help-seeking and perceptions and beliefs around mental health and mental health treatment. Third, MEPS does not include information about English language proficiency, a substantial and well documented barrier to care which is likely to impede the ability of some Asian Americans to utilize mental healthcare. We adjust for US birth and time in the USA, two variables which may serve as proxies for English language ability. Fourth, MEPS examines cultural factors, such as stigma or alternative conceptions of mental health symptoms, which may impact healthcare-related behavior. However, this underscores our main findings—even when Asian Americans report symptoms which align with common presentations of mental illness, they are less likely to receive care. Fifth, MEPS excludes homeless and incarcerated individuals, languages other than English/Spanish, and does not accurately measure undocumented immigrant status (which would be a large barrier to access). Thus, our findings are conservative estimates. Sixth, we used the PHQ-2 and K-6 scales to define populations with elevated risk for mental illness opposed to clinical evaluations. Nonetheless, these scales have well-validated psychometric properties for use for diverse patient populations and have been used for similar purposes in previous research. Given these limitations and the importance of this topic, future research should aim to examine use among additional Asian American subgroups while incorporating data addressing factors which could not be examined in this analysis.

Conclusion

Disparities in mental healthcare access among Asian Americans, including Asian Indians, Chinese, Filipinos, and Other Asians/Pacific Islanders relative to White Americans persist, even among those with elevated risk for mental illness. Stakeholders should recognize the importance of utilizing disaggregated data for this heterogeneous population whenever possible. Future interventions should address treatment disparities impacting the Asian American population with a focus on eliminating barriers to accessible, culturally competent care.