Perception of Need and Receipt of Mental Health Treatment: A Three-Group Comparison of Young Adults With Psychological Distress

This study combined public-use data from six years (2008–2013) of the National Survey on Drug Use and Health (NSDUH), which applies a multistage, stratified random sampling design. The study included participants ages 18 to 25 with serious psychological distress (N=19,775), 18% of all individuals in this age group who were surveyed in the NSDUH (N=112,665). Most study participants were white (62.1%) and female (61.3%). Gender was disproportionately female compared with the overall U.S. population, but the distribution of racial-ethnic minority groups was similar to that in the U.S. population. [A table summarizing other characteristics of the sample is included in an online data supplement to this report.] Study procedures were approved by the University of Houston Institutional Review Board.

Psychological distress was assessed with the Kessler-6. Respondents rated frequency of distress from 0, never, to 4, very often, for six symptoms over the past year, for a possible total score range of 0–24. Scores of 13 or higher were considered indicative of serious psychological distress on the basis of prior work (13).

Our primary dependent variable was perceived need/treatment group. Participants were divided into three mutually exclusive groups: no perceived need and no mental health treatment, perceived need and no mental health treatment (unmet need), and received mental health treatment. Those who had not received mental health treatment and did not state that there had been a time when they needed treatment were classified in the group no perceived need and no treatment. We excluded anyone who reported receipt of alternative treatment for a psychological problem (for example, from a chiropractor) (N=967), because this indicated awareness of a problem. For the unmet-need group, we used the question, “During the past 12 months, was there any time when you needed mental health treatment or counseling for yourself but didn't get it?” Those who answered affirmatively and had not received treatment were assigned to the second group. Those who had received any outpatient treatment, inpatient treatment, or psychiatric medications in the past year were assigned to the third group.

Independent variables were grouped as demographic, access related, and need related. Demographic variables included gender, marital status, race-ethnicity (white, African American, Asian, Hispanic, and other), and education (collapsed into high school or less and some college or more). Access-related variables included having health insurance, income, military service, employment, current enrollment in school, arrested in the past year, county size (large metro, small metro, and nonmetro), moving six or more times in the past five years, and treatment for substance use in the past year. Need-related variables included a substance use disorder (either abuse or dependence), self-reported health (excellent or very good versus good, fair, or poor), and score on the Kessler-6.

Multinomial logistic regression was used to assess predictors of membership in each group compared with each of the others. Survey weights were used to generalize to U.S. population estimates for age, gender, and race-ethnicity. Analyses used imputed data provided in the public-use data sets and settings to adjust for survey design and weighting in Stata 12. Given the large sample and number of comparisons, we used the Bonferroni correction to adjust statistical significance to p<.002 (.05/23 comparisons).

In this sample of young adults who met criteria for psychological distress, half (51.0%, N=9,997) gave no indication that they perceived a need for mental health treatment. One-third (33.7%, N=6,673) had received treatment in the prior year, and 15.7% (N=3,105) reported unmet need. Predictors of membership in each of the groups are presented in Table 1. Significant differences in race-ethnicity and gender were noted between the group with no perceived need for treatment and the group with unmet need. African-American, Hispanic, or Asian respondents were more likely than white respondents to be in the former group than in either of the other groups. Males were also more likely to be in the group with no perceived need than in either of the other groups. Respondents with higher need (as indicated by poor health status, presence of a substance use disorder, and high Kessler-6 scores) were also significantly less likely to be in the group with no perceived need than in either of the other groups.

Respondents who received mental health treatment were less likely to be black or Hispanic, less likely to be uninsured, and less likely to have only a high school education, compared with either of the other groups (Table 1). Those with substance use disorders were more likely to be in the group with unmet need than in the group that received treatment; however, receipt of substance abuse treatment was a predictor of being in the group that received mental health treatment versus the group with unmet need. No gender differences were found between the group with unmet need and the group that received treatment.

This study of a large, nationally representative sample adds to the literature by examining differences between perception of need and receipt of mental health treatment among young adults with psychological distress. Our findings suggest that rates of treatment among young adults are partly related to unacknowledged need for treatment. More than half of our sample did not perceive a need for treatment. This finding aligns with literature on the transitional period in which young people are exploring their identities and may be hesitant to acknowledge a need for formal treatment (9). Interventions seeking to increase treatment rates for young adults should be targeted to reach those who do not acknowledge a need for treatment. Broad-based education framed around recognizing symptoms and destigmatizing treatment might facilitate earlier problem recognition and treatment seeking. Information that helps individuals better understand when to seek mental health treatment and what to expect from treatment may increase service use.

Although perceptual factors were clearly important in the study, access and need factors also played a role. Receipt of substance abuse treatment may facilitate access because it was significantly related to receipt of mental health treatment; however, having a substance use disorder predicted membership in the group with unmet need. Identification of comorbid problems is critical in this developmental period when substance use peaks. Having insurance was also significantly related to treatment receipt, highlighting lack of health insurance as an ongoing barrier to treatment (14). Although the Affordable Care Act (ACA) has begun to expand coverage by extending dependent care coverage until age 26 for some individuals, offering subsidies for those up to 400% of the federal poverty level, and expanding Medicaid coverage in some states, some of the most vulnerable young adults still have limited access to care. Youths who live where Medicaid has not been expanded and who have low-wage employment or parents who cannot cover them are likely to remain uninsured (14) and to use fewer mental health services. These young people are also likely to have lower levels of education, a factor that was negatively associated with mental health service use in this study. Attending college may provide increased knowledge and awareness of mental health and ease access to services through campus resources. Special efforts are needed to support young adults who have not attended college, remain uninsured, and have comorbid substance use disorders.

Demographic differences among the three groups provide information to inform tailoring of interventions. We found that men were less likely than women to perceive a need for mental health treatment, but once need was perceived, men were equally likely to receive treatment. Providing education tailored to men about understanding and acknowledging symptoms appears the most beneficial approach to increasing their service use. Consistent with previous studies, whites were more likely than both Hispanics and African Americans to receive treatment (1,2). However, we also found racial-ethnic differences in perception of need. This indicates that racial disparities in mental health treatment may be attributed both to access factors and to differences in the construction of psychological distress as a treatment-worthy problem.

Our study had several limitations. First, “no perceived need” was an indirect measure constructed on the basis of a lack of an affirmative response to questions. Second, we used a measure of psychological distress to indicate need, which is imperfect without formal assessment of diagnoses and impairments. The group with no perceived need had significantly lower scores on the Kessler-6 and reported less substance use and better health than the other two groups. This could indicate that the need for treatment is actually lower in this group or that the symptoms, while distressing, have not yet risen to the point of motivating treatment seeking. Finally, the dependent care extension of the ACA took effect midway through the six years of the study. We controlled for year in the analyses, but the ACA may have changed access in unmeasured ways that had an impact on our results.

More than half of young adults with assessed psychological distress did not perceive a need for treatment. These young adults were more likely to be male, to be from racial-ethnic minority groups, and to have a high school education or less. Efforts to increase service utilization among young adults should provide targeted education about mental health symptoms and treatments with these subgroups in mind. In addition, easing barriers, such as lack of insurance, continues to be an important target for ensuring access to treatment once need is perceived.

The authors appreciate statistical and editorial support from Paul Swank, Ph.D., and Jodi Berger-Cardoso, Ph.D.