Impact of social determinants of health and individual characteristics on substance use disorder treatment utilization among emerging adults aged 18-25

Abstract

Background:

Emerging adulthood (typically ages 18-25) is a developmental period associated with increased likelihood of developing substance use disorders (SUDs) resulting in life-altering negative consequences. While SUD treatment can be effective and diminish substance-related consequences, very few emerging adults with SUDs actually receive treatment and research is needed to identify potential determinants of receiving care. We examined associations between individual factors and social determinants of health (SDOH) with SUD treatment service utilization to inform treatment engagement strategies for emerging adults.

Methods:

Using pooled cross-sectional data from the 2015-2019 years of the National Survey on Drug Use and Health, we evaluated past-year treatment utilization among emerging adults (i.e., respondents aged 18-25 years-old) with a past-year SUD (N = 12,594). Logistic regression analyses evaluated sex, race, education level, lifetime arrest history and age of alcohol initiation in addition to mental illness, employment, food stamp utilization, poverty status, and frequency of housing relocation.

Results:

Past-year mental illness, Non-Hispanic White identity, not having college degree, lifetime arrest history, part-time employment or unemployment, and relocating for housing more than 2 times in the past year were significantly associated with greater likelihood of treatment utilization in adjusted analyses.

Conclusions:

Findings from this nationally representative survey of emerging adults highlight the possible contribution of some SDOH associated with social disadvantage and/or criminal justice involvement on SUD treatment utilization. These results may guide providers in tailoring service delivery mode, community organizations to target populations, and implementation strategies that are effective in reducing SUD treatment service utilization inequity.

Introduction

The transition from adolescence to adulthood, often referred to as emerging adulthood (ages 18-25), is associated with higher propensity for risk-taking behaviors (Arnett, 2005; Arnett, 2000; Institute of Medicine and National Research Council, 2011; World Health Organization, 2018) compared to childhood and established adulthood. Alcohol and Other Drug (AOD) use peaks during this developmental period (McCabe et al., 2019; Seth et al., 2018). For example, national data from the US in 2021 showed that emerging adults’ past-month binge drinking prevalence (29.2%) and past-year illicit drug use (35.4%) were higher compared to adolescents ages 12-17 (3.8%, 10.5%) and established adults ages 26 and older (22.4%, 17.2%) (Center for Behavioral Health Statistics (CBHS), 2022). Similarly, the percentage of people who were classified as having past-year alcohol or drug abuse or dependence (i.e., substance use disorders [SUDs]) was highest (25.6%) among emerging adults versus these other ages (i.e., 16.1% for ages 26+ and 8.5% for ages 12-17; CBHS, 2022).

AOD use during emerging adulthood is associated with numerous social and physiological life-altering negative consequences. These include injuries and deaths related to impaired driving (Wu et al., 2017), unwanted/unprotected sexual activities (SAMHSA, 2015), SUDs later in life (Grant et al., 2001; Perkonigg et al., 2006), and higher rates of engagement in crime (Rosenfeld et al., 2012; Turner et al., 2004). Among emerging adults with SUDs, SUD treatment can be effective and can diminish the longer-term negative consequences of substance use (Evans et al., 2013). However, only approximately 8% of emerging adults who could benefit from SUD treatment received these services within the past year (SAMHSA, 2020).

Prior research using nationally representative adult samples has examined specific individual characteristics associated with receiving SUD treatment. Results of these studies have been mixed. Some factors that have been consistently linked to higher SUD treatment utilization included male sex (Cohen et al., 2007; Haughwout et al., 2016; Wu et al., 2003), comorbid mental health issues (Edlund et al., 2012; Grella et al., 2009; Kerridge et al., 2017; Wu et al., 2003, 2016), and older age (Cohen et al., 2007; Grella et al., 2009; Haughwout et al., 2016). Findings related to race/ethnicity, and financial status have been more inconsistent. Some studies have found those identifying with historically marginalized races/ethnicities, such as Black and Hispanic individuals, to be more likely to receive SUD treatment (Cohen et al., 2007; Grella et al., 2009) whereas others have found that Non-Hispanic White individuals were more likely to receive treatment (Hatzenbuehler et al., 2008; Wu et al., 2011, 2016). Similarly, in terms of financial status, results conflicted (Cohen et al., 2007; Kerridge et al., 2017). The variability in findings from these nationally representative surveys may be due to the range of age groups included (i.e., emerging adults, older adults).

Although limited in quantity and scope, there are some studies that have focused on emerging adults’ SUD treatment utilization. In one study, college students were less likely to seek treatment than their Non-student peers (Wu et al., 2007) whereas another study reported childhood maltreatment being associated with a greater likelihood of accessing SUD treatment (Goldstein et al., 2013). In addition, a nationwide survey of students enrolled in Collegiate Recovery Programs who received counseling reported that most of the students had used multiple substances, had high SUD severity, criminal justice involvement, co-occurring addictions, and homelessness (Laudet, 2015). Notably, it was reported that individuals with an earlier age for onset of drinking is associated with alcohol and substance abuse during emerging adulthood implying the early onset age of drinking may influence emerging adults’ treatment needs (Ohannessian et al., 2015).

The social determinants of health (SDOH) framework which emphasizes the conditions in the environments where people are situated that affect health related outcomes and risks can provide guidance in understanding and exploring factors that influence treatment utilization disparities. The Healthy People national initiative declared SDOH a priority area, highlighting its important in improving health and reducing health disparities (Social Determinants of Health - Healthy People 2030 | Health.Gov, n.d.). Aligned with these goals, scholars in the healthcare services field have advocated for a multidimensional conceptualization (i.e., person in environment perspective) to enhance the understanding around health service utilization (Levesque et al., 2013). More specifically for SUD treatment utilization studies, economic resources (Ilgen et al., 2011; Kerridge et al., 2017), health insurance (Wu et al., 2016), and different social circumstances such as criminal justice involvement (Haughwout et al., 2016) have been examined. However, these prior efforts have typically focused either on adolescents (ages 12-17) or adults without a unique focus on emerging adults despite potential differences in treatment-seeking during this developmental period. For example, emerging adults’ treatment readiness, perceived need for treatment, motivation to change have consistently been reported as lower than adolescents’ (Grant, 1995; Satre et al., 2003; Sinha et al., 2003). Furthermore, emerging adults have greater social network support for AOD use (Andrews et al., 2002; Delucchi et al., 2008) and experience multiple relocations in housing, changing relationships, and frequent unemployment compared to established adults (Arnett & Tanner, 2006; Brown et al., 2008) making it more difficult for emerging adults to receive SUD treatment.

In addition to a lack of focus on emerging adults specifically, many prior studies were conducted using data collected before or partly during the enactment (i.e., year 2010) of the Affordable Care Act (ACA; Pub L No. 111–148) which dramatically changed the landscape of SUD treatment access through major coverage expansions (Abraham et al., 2017; Humphreys & Frank, 2014). The present study extends prior work by using the largest annual survey on substance use in the United States (US), the National Survey on Drug Use and Health (NSDUH), and pooling cross-sectional data for years 2015-2019 to examine the factors associated with SUD treatment utilization among emerging adults with SUDs. The investigation took a person in environment perspective where the relative contribution of SDOH in addition to previously identified individual level risk factors (i.e., biological sex, race, age of alcohol initiation, and any mental illness in the past year) were examined as they relate to SUD treatment service utilization. The knowledge on SDOH factors in conjunction with individual characteristics that affect treatment utilization among emerging adults with SUDs could inform efforts to reduce disparities in treatment for this population.

Based on the findings of previous research summarized above, we hypothesize that some of the individual characteristics including comorbid mental illness, early age of alcohol initiation, and male sex would be linked with higher SUD treatment service utilization. Some of the SDOH factors that we hypothesize to be related to higher SUD treatment service utilization are no college education, lifetime arrest history, unstable housing, and health insurance coverage. There was not enough existing literature providing information around how race, employment, and financial status are associated with SUD treatment service utilization among emerging adults.

Methods

Study Design

We analyzed data from the NSDUH from 2015-2019. The NSDUH is a nationally representative cross-sectional survey of household respondents in the United States and the District of Columbia. The NSDUH is administered by Substance Abuse and Mental Health Services Administration to provide national estimates on mental health and AOD use. All participants provided informed consent to the NSDUH methodology team, and all data were deidentified and made publicly available. Additional information on NSDUH and its methodology has been previously reported (CBHS, 2019). The current study was exempt from the local institution’s Institutional Review Board.

Sample

The NSDUH used a multi-stage probability stratified sampling technique to collect cross-sectional data annually from respondents aged 12 years or older. The current analysis focused on the emerging adult sub-sample (i.e., respondents who are 18-25 years-old). Within this emerging adult sample (N = 84,469), respondents who had a past-year SUD (i.e., substance abuse or dependence; N = 12,594) related to the following substances were identified: alcohol, cannabis, hallucinogen, inhalants, methamphetamine, tranquilizers, cocaine, heroin, pain relievers, stimulants, or sedatives. The NSDUH SUD questions were based on the criteria in the American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders, 4th edition. Respondents were defined as having abuse or dependence (i.e., SUD) if they self-reported a positive response to abuse and dependence criteria consistent with DSM-IV (detailed criteria can be found in Appendix A).

Measures

SUD Treatment Utilization

NSDUH’s item assessing past-year treatment utilization (our dependent measure herein) was as follows: “During the past 12 months, that is since [DATEFILL] have you received treatment or counseling for your use of alcohol or any drug, not counting cigarettes?” The respondents were instructed to report treatment or counseling designed to help then reduce or stop their alcohol or drug use including detoxification and any other treatment for medical problems associated with their alcohol or drug use.

Individual Characteristics

Respondents’ past-year mental illness, age of alcohol initiation, biological sex, and race were included in the analysis. The past-year mental illness was a binary indicator for Any Mental Illness (AMI) based on the 2012 revised predicted probability of mental illness where AMI was defined as having Serious, Moderate, or Mild Mental Illness and does not include SUDs. Age of alcohol initiation was a continuous variable and used responses from a question asking respondents how old participants were in year when they drank for the first time. Sex was a binary variable with male/female options. Race was a categorical variable with 4 options including Non-Hispanic White, Non-Hispanic Black/African American, Non-Hispanic Others, and Hispanic.

SDOH Characteristics

SDOH identified by the Healthy People 2030 framework (Health.Gov., 2023) corresponding to the available measurements within NSDUH data were included in the model. All analogous variables identified within NSDUH data are listed in Table 1. Education was a binary variable indicating those with any college education and those less than high school or high school graduate. Lifetime arrest history was also a binary variable. Past-year employment had three options including employed full-time, employed part-time, and unemployed. Number of times moved in the past year (a marker of housing instability) had three options including zero, one time, and more than two times. Past-year poverty status was also coded with three categories including those with income more than two times the federal poverty threshold, living in poverty, and income up to two times the federal poverty threshold. Past-year health insurance coverage was a binary variable.

Table 1.

Proxy variables within NSDUH data for SDOH identified by Healthy People 2030.

SDOH Identified by Healthy People 2030	Proxy NSDUH Data Measures
Education Access and Quality
Enrollment in Higher Education	Highest education achieved
Economic Stability
Employment	Past-year employment
Housing instability	Number of moves in the past year
Poverty	Past-year poverty status
Health Care Access and Quality
Access to Health Services	Past-year health insurance coverage
Social and Community Context
Incarceration	Lifetime arrest history

Data Analysis

We focused our analysis only on emerging adults who had SUD and conducted statistical analyses using complex sample analysis with IBM SPSS software, version 28.0. This system adjusts standard errors of estimates for complex survey sampling design effects including clustered multistage data. All analyses were weighted using final person-level analysis weights stipulated by NSDUH to provide nationally representative estimates. This person-level analysis weights accounts for either a selection probability at a selection stage or an adjustment factor adjusting for nonresponse, coverage, or extreme weights. See the NSDUH Methodological Resource Book (CBHS, 2019). To examine the individual and SDOH predictors of SUD treatment utilization, logistic regression analysis was used. Specifically, individual characteristics (i.e., gender, race, age of alcohol initiation, and any mental illness past year) were included along with SDOH factors (i.e., education, employment, arrest history, number of times moved in the past year, poverty status, and health insurance status) in logistic regression model evaluating SUD treatment utilization.

Results

The total number of emerging adults ages 18-25 within the NSDUH 2015-2019 data was 84,469. Overall, 14.9% of respondents had a past-year SUD. Individual characteristics and SDOH factors for the emerging adults with SUDs by past-year substance use treatment utilization are summarized in Table 2.

Table 2.

Weighted Characteristics of emerging adults (EAs) from NSDUH 2015-2019

	SUDs (N = 12,594, 14.9%)*
	Total	Treatment n = 1020 8.1%	No Treatment n = 11574 91.9%
Past-year SUD	100.0%	100%	100%
Past-year mental illness	44.2%	57.3%	43.0%
Age of alcohol initiation	16.09 (.368)	14.43 (.121)	16.21 (.398)
Biological sex
Male	57.6%	60.8%	57.3%
Female	42.4%	39.2%	42.7%
Race
Non-Hispanic White	60.4%	65.1%	59.9%
Non-Hispanic Black and African American	11.4%	9.1%	11.6%
Non-Hispanic Others (Native/HA/PI/Asian/Multi)	8.1%	8.0%	8.1%
Hispanic	20.1%	17.8%	20.3%
Highest education achieved
Less than high school or graduated high school	44.2%	58.1%	42.9%
Any college	55.8%	41.9%	57.1%
Lifetime arrest history	28.1%	65.9%	24.8%
Past-year employment
Employed full-time	45.6%	39.1%	46.2%
Employed part-time	24.7%	19.6%	25.2%
Unemployed	29.7%	41.3%	28.6%
Number of times moved in past year
0 time	47.1%	42.6%	47.5%
1 time	29.4%	24.2%	29.8%
2+ times	23.5%	33.2%	22.7%
Past-year poverty status
Living in poverty	22.4%	21.8%	22.5%
Income up to 2X federal poverty threshold	52.6%	49.5%	52.9%
Income more than 2X federal poverty threshold	24.9%	28.7%	24.6%
Past-year health insurance coverage	85.5%	85.4%	85.5%

Note: Bolded cells indicate a significant difference between the Treatment and No Treatment groups at p < .001. Significance is based on the adjusted F and its degrees of freedom where F is a variant of the second-order Rao-Scott adjusted chi-square statistic.

^*Alcohol Use Disorder (AUD) only (n =6,558, 52 .1%)

Drug Use Disorder (DUD) only (n = 3,833, 30.4%)

AUD+DUD (n = 2203, 17.5%)

The logistic regression model was statistically significant, χ²(36) = 28.977, p < .001, Nagelkerke’s R² = 0.17. See Table 3 for odds ratios and results.

Table 3.

Logistic regression results evaluating likelihood of treatment utilization based on individual characteristics and SDOH factors

	Dependent Variable [Ref: No Treatment]	p	Odds Ratio	95% C.I.Pair-Wise
				Lower	Upper
Individual Characteristics	Past-year mental illness	<.001	1.811	1.438	2.280
	Age of alcohol initiation	<.001	.937	.905	.970
	Biological sex [Ref: Female]	.714
	Male		.960	.771	1.197
	Race [Ref: Non-Hispanic White]	.005
	Non-Hisp Black/African American		.595	.416	.850
	Non-Hisp Othersa		.871	.570	1.332
	Hispanic		.710	.544	.927
SDOH Factors	Education [Ref: Any college]	.011
	Less than high school or high school graduate		1.364	1.078	1.725
	Lifetime arrest history	.000	5.658	4.567	7.008
	Past-year employment [Ref: Employed full-time)	.006
	Employed part-time		1.159	.830	1.617
	Unemployed		1.478	1.170	1.867
	Number of times moved past year [Ref: 0 times]	.006
	1 time		.878	.695	1.111
	2+ times		1.392	1.037	1.868
	Past-year poverty status [Ref: more than 2X federal poverty threshold]	.530
	Living in poverty		.869	.646	1.169
	Income up to 2X federal poverty threshold		.864	.662	1.128
	Past-year health insurance coverage	.108	1.305	.941	1.808

Note:

Bolded cells represent results that were statistically significant at p < .01.

95% CIs correspond to pair-wise comparison results to show which categories were specifically significant in relation to the reference category of the variable.

^a.Non-Hispanic Others (Native/HA/PI/Asian/Multi)

Ref. Reference group for all predictors were chosen based on our hypotheses where the represented options of the categorical variables are the ones hypothesized to have increased likelihood of treatment utilization.

According to our model, having past-year mental illness, younger age of alcohol initiation, identifying as Non-Hispanic White, and not having college experience were statistically significantly associated with greater likelihood of treatment utilization in the past year. Specifically, emerging adults who began drinking at younger ages were more likely to get SUD treatment compared to those who started drinking at an older age. Arrest history, being employed part-time or unemployed, and relocating for housing more than 2 times in the past year were statistically significant SDOH factors associated with higher likelihood of utilizing SUD treatment.

Discussion

In the present study we examined the associations between several individual and SDOH factors with utilization of SUD treatment among a nationally representative sample of emerging adults in the U.S. The sample was restricted to those with past-year SUDs between 2015-2019. Overall, approximately 8% of emerging adults with a past-year SUD reported receiving SUD treatment. Given that fewer than one in ten emerging adults with SUDs receive SUD treatment, clear opportunities exist to expand treatment utilization in this population with consideration of factors examined herein.

Adjusted analyses revealed that emerging adults with a past-year mental illness, younger age of alcohol use onset, and Non-Hispanic White ethnoracial identities were significantly more likely to have received treatment for SUDs. These findings align with existing literature reporting greater SUD treatment utilization among those with comorbid mental health conditions (Edlund et al., 2012; Grella et al., 2009; Kerridge et al., 2017; Wu et al., 2003, 2016), and younger age of initiation of alcohol use (Ohannessian et al., 2015) among the general adult population. While findings related to race/ethnicity have been more inconsistent, results of the current study align with some general adult population studies that have found Non-Hispanic white individuals were more likely to receive treatment (Hatzenbuehler et al., 2008; Wu et al., 2011, 2016). The present analyses did not find any statistically significant differences in SUD treatment utilization by biological sex. This finding contrasted with previous reports citing sex differences among general adult population for their SUDs treatment utilization (Cohen et al., 2007; Haughwout et al., 2016; Wu et al., 2003) and indicate that sex differences may be less pronounced in emerging adults than the entire adult population.

Among seven SDOH factors included in the model, four factors were significantly associated with treatment utilization. Compared to those with college degree or higher education, emerging adults with lower levels of educational attainment had approximately double the odds of utilizing the treatment. Those who were employed part-time and unemployed had higher odds of treatment receipt compared to those who were employed full-time. Findings also demonstrate that frequent relocations for housing were related to higher odds of treatment utilization. These results were inconsistent with the emerging adulthood literature which has previously indicated that multiple relocations in housing and frequent unemployment may make it more difficult for emerging adults to receive SUD treatment (Arnett & Tanner, 2006; Brown et al., 2008). Taken together, the findings of the present analyses indicate that factors that are typically linked to social disadvantage tend to be associated with greater likelihood of receiving SUD treatment in this population. Prior research in adults has found that many individuals in SUD treatment report that they were seeking care because of external pressures, often referred to as extrinsic motivation (Klag et al., 2010; Wild et al., 2006), and these may have been at least partially driving the small portion of emerging adults receiving care in the present sample into services.

It is worthwhile to highlight the arrest history resulted in almost 5.7 times higher likelihood of using SUD treatment. This may reflect the possibility that the criminal justice-involved emerging adults are more frequently evaluated and formally diagnosed with SUD as part of court proceedings, thus via such assessments may more readily identify SUD symptoms. However, studies have also demonstrated a large unmet need for SUD treatment among current and formerly incarcerated individuals (Nam et al., 2016; National Center on Addiction and Substance Abuse, 2010; Rowell-Cunsolo & Bellerose, 2021). Future studies should examine the manifestation of unmet treatment needs despite higher treatment utilization for this particular population. Interestingly, poverty status and health insurance status were not related to treatment utilization in the present analyses. These null findings are nonetheless still important and warrant qualitative research that will help elucidate the process through which these important SDOH contribute to and potentially increase SUD treatment utilization among emerging adults.

Some limitations of the study are important to note. First, data collected after 2019 was not included as the study design was drastically modified due to COVID-19 pandemic. Institutionalized individuals are also excluded from the NSDUH by the study design. In addition, although the NSDUH did allow for an initial examination of how some SDOH related to SUD treatment utilization, more detailed information on all five domains of SDOH (i.e., economic stability, education access and quality, health care access and quality, neighborhood and built environment, and social and community context) would help to deepen the understanding of the relative importance of each SDOH areas. Further, the analytic sample only included individuals with past-year SUD identified based on self-reported survey items within the NSDUH and our results may be confounded by the level of self-awareness of SUD symptoms. The analysis also did not account for severity of the SUD. Accordingly, it may be that those with no college education and unemployed individuals are more likely to have received treatment due to their more severe SUD. Finally, the analyses were cross-sectional and it is possible that some of the factors examined as potential drivers of SUD treatment utilization occurred following or as a result of SUD treatment use.

Conclusions

The current study provides an important contribution to the literature by identifying both individual characteristics and SDOH factors related to utilization of SUD treatment among emerging adults. Specific factors within the SDOH domains were categorized to be consistent with the Healthy People 2030 framework. In particular, factors associated with social disadvantage and/or criminal justice involvement increased the likelihood of SUD treatment utilization in those with SUDs. Service delivery modes should be considered that are user-friendly and barrier/stigma free for emerging adults with these social disadvantage and/or criminal justice involvement in order to increase their utilization. Although additional research addressing limitations of this study are warranted, community organizations should be cognizant that emerging adults in college or those who have education beyond high school, full-time employees, and those with low-income were less likely to utilize treatment. Given their lack of utilization, outreach strategies to meet emerging adults’ where they are may be particularly impactful. Future research should consider other SDOH and longitudinal associations between these factors and service utilization by emerging adults.