Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Sep 26.
Published before final editing as: Subst Use Addctn J. 2026 Aug 9:29767342261471776. doi: 10.1177/29767342261471776

Disability-Related Differences in Substance Use Among U.S. Young Adults

Jessica Williams a, Xiao Li a, Carla J Berg b, Hannah Szlyk a, Erin Kasson a, Patricia A Cavazos-Rehg a
PMCID: PMC13613204  NIHMSID: NIHMS2208287  PMID: 42571633

Abstract

Background:

People with (vs. without) disabilities may be at greater risk for substance use, but limited research has explored use of various substances among people with disabilities. This study examined differences in substance use among U.S. young adults with vs. without disabilities.

Methods:

Survey data was derived from an internet-based, purposive national sample of U.S. young adults (aged 18–34) (N=3,458; Mage=26.4; 61.2% female, 63.2% White). Multivariable regression analyses examined disability status, subtype (hearing, vision, cognitive, mobility), and number of disabling limitations in relation to: past 30-day cannabis and tobacco use frequency (ordinal logistic regression, adjusted odds ratios (aOR)); past 30-day binge drinking (negative-binomial regression, adjusted incidence rate ratio (aIRR)); and past 6-month non-medical stimulant and opioid use (logistic regression, aOR).

Results:

One-third of participants (30.2%) reported having at least one disability. 40.5%, 37.8%, and 33.1% participants reported past 30-day cannabis use, tobacco use, and binge drinking (respectively). 8.7% and 5.3% of participants reported past 6-month non-medical stimulant and opioid use, respectively Participants with (vs. without) disability had greater odds of more frequent cannabis use (aOR: 1.44, 95%CI: 1.23–1.69), more days of tobacco use (1.61, 1.33–1.93), and nearly three times higher rates of both non-medical use of stimulants (3.02, 2.23–4.09) and opioids (3.00, 2.11–4.28). Mobility/complex activity limitations were associated with more days of binge drinking (1.55, 1.08–2.21). Compared to participants without disabilities, a greater number of disability subtypes was associated with higher likelihood of using all substances studied, except for binge drinking.

Conclusions:

Young adult substance use patterns differ by disability status, disability subtype, and number of disabling limitations, as well as by specific substances. Future research should prioritize tailored prevention and treatment strategies to address substance use within this underserved population with disabilities.

Keywords: persons with disabilities, young adult, substance-related disorders

INTRODUCTION

Substance use continues as a major public health challenge in the United States, especially among young adults.1–3 This life stage often marks a person’s first exposure to alcohol, cannabis, and illicit drugs. Coinciding with increasing state legalization, young adults are using cannabis at record levels,4 while tobacco and alcohol use remain common,5 and stimulant and opioid use propel addiction and premature overdose deaths in this group.6 Substance use beginning in young adulthood often escalates over time, increasing risk for substance use disorders (SUD) and adverse long-term health and psychosocial outcomes.7,8

National surveillance systems such as the National Survey on Drug Use and Health (NSDUH) offer nationally representative estimates of substance use.9 Prior analyses show elevated substance use rates among young adults with disabilities, particularly for cannabis, tobacco, and alcohol.10–12 However, high-risk substance use, like non-medical use of stimulants and opioids, remains understudied among people with disabilities (PWD). There is also a need to move beyond prevalence estimates to examine substance use within discrete, clinically relevant timeframes.

The present study addresses this gap using a national, diverse convenience sample of U.S. young adults (aged 18–34) from a longitudinal cohort.13 Disability status was assessed using standardized measures like those in NSDUH,14 allowing granular analyses of disability subtype and number of limitations in relation to five substances. We address two questions: (1) Are young adults with (vs. without) disabilities more likely to use cannabis or tobacco, binge drink, or report non-medical use of opioids or stimulants? (2) How do these associations differ by substance or by type and number of disabilities?

METHODS

Data Source and Analytic Sample

This cross-sectional analysis used data from Wave 3 (June–November 2024) of the Cannabis Regulation, Marketing, and Appeal (CARMA) study, a longitudinal cohort study examining the impact of non-medical cannabis retail access, marketing, and use. Eligibility criteria included age 18–34 years, U.S. residency, and English fluency. Purposive, quota-based, online convenience sampling recruited a cohort of 50% past 30-day cannabis users, 50% males and females, and 40% racial/ethnic minorities. Our design captured higher substance use rates in young adults, supporting our aim of examining associations across disability characteristics and substance use patterns. Facebook ads were deployed nationwide from June-November 2023; interested participants were screened for eligibility and prompted to complete the baseline (Wave 1) survey. Every six months, they were asked to complete another survey wave. Further description of the CARMA study design and recruitment procedures is available in prior work.13 The study protocol was approved by the George Washington University Institutional Review Board (IRB# NCR224124).

Measures

All measures were collected in Wave 3, except for socio-demographics (Wave 1).

Outcome Measures of Substance Use

Cannabis use was assessed by asking the number of days of use in the past 30-days (0–30). The survey defined cannabis-related products as: “marijuana (i.e., cannabis, pot, or weed), including all forms of the plant and its preparations—such as dried herb, edibles, oils, hash, kief, concentrates, cannabis-infused beverages, tinctures, lotions, and other cannabis products. Do not include hemp-derived cannabinoids, like Delta-8.” Participants were categorized into four groups: non-use (0 days), infrequent (1–5), frequent (6–29), and daily (30).15

Tobacco product use was assessed by asking the number of days of use in the past 30-days (0–30) of: traditional cigarettes, large/little cigars, e-cigarettes (or other electronic nicotine delivery systems that vape nicotine), hookah (tobacco waterpipe/narghile), smokeless tobacco (e.g., chew, snus), tobacco-free nicotine pouches. If a participant used more than one tobacco product, the maximum number of days reported across products was used as the outcome. Participants were categorized into four groups: 0, 1–9, 10–24, and >25 days.16

Binge drinking was assessed by asking: “In the past 30-days, on how many days did you have 4 or more alcoholic drinks (if female) or 5 or more drinks (if male) on a single occasion?” The outcome was assessed as a count variable, ranging from 0 to 30.

Non-medical use of stimulants was assessed separately for prescription (e.g., Ritalin, Concerta, Adderall) and non-prescription stimulants (e.g., cocaine, methamphetamine). For each, participants were asked: “In the past 6 months, how many days did you use [stimulant]?” Answer choices included: 0, 1–30, 31–60, 61–90, 91–120, 121–150, and 151–180 days; responses were dichotomized as “any” vs. “no use.” Among those reporting use of prescription stimulants, a follow-up item asked if any of the stimulants were prescribed for the participant (none/some/all). Non-medical stimulant use was defined as using any non-prescription or any prescription stimulants that were not prescribed to them. Participants reporting only prescribed stimulant use were excluded (n=215).

Non-medical use of opioids was assessed separately for prescription (e.g., oxycodone, morphine, codeine) and non-prescription opioids (e.g., heroin). For each, participants were asked: “In the past 6 months, how many days did you use [opioid]?” Answer choices included: 0, 1–30, 31–60, 61–90, 91–120, 121–150, and 151–180 days; responses were dichotomized as “any” vs. “no use.” Among those reporting use of prescription opioids, a follow-up item asked if any of the opioids used were prescribed for the participant (none/some/all). Non-medical opioid use was defined as using any non-prescription or any prescription opioids that were not prescribed to them. Participants reporting only prescribed opioid use were excluded (n=149).

Disability Measure

Participants were asked about disabilities and six specific limitations: “Please indicate if you have any of the following: (Check all that apply): (1) A disability, defined by the Americans with Disabilities Act (ADA) as a physical or mental impairment that substantially limits one or more life activities; (2) deafness or serious difficulty hearing; (3) blindness or serious difficulty seeing, even when wearing glasses; (4) difficulty walking or climbing stairs; (5) difficulty dressing or bathing; (6) difficulty doing errands alone, such as visiting a doctor’s office or shopping, due to a physical, mental, or emotional condition; and (7) serious difficulty concentrating, remembering, or making decisions due to a physical, mental, or emotional condition.”17 Participants who endorsed the ADA-defined disability item (1) or any of the six disabling limitation items (2–7) were classified as having any disability (vs. none). Two indicators of disability status were derived: (1) disability subtype, in mutually exclusive categories (i.e., none vs. [limitation] only vs. ≥2 limitations); and (2) number of disabling limitations (0, 1, 2, or ≥3).18,19 Given the low prevalence of “mobility limitation only” (<0.5%), individuals with “mobility limitation only” and “complex activity limitation only” were combined into one category, following prior research.20 Participants who selected the ADA disability item without designating subtype(s) were included in overall disability counts, though their subtype was considered unavailable data and they were excluded from the subtype-specific regression models.

Covariates (Socio-Demographics, Cannabis Legalization, and Mental Health Symptoms)

The survey assessed age, sex at birth, ethnicity, race, educational attainment, employment status, community type (e.g., rural, urban), relationship status, and parenthood. Cannabis legalization status was operationalized as residing in a state with legalized non-medical adult cannabis use. Depressive symptoms were measured using the Patient Health Questionnaire-8 (PHQ-8), which excludes the ninth item regarding suicidal ideation.21 Anxiety symptoms were measured using the Generalized Anxiety Disorder-7 (GAD-7). Cutoffs at 10 and 14 were used to categorize severity: none/mild (<10), moderate (10–14), and severe (>14).22

Analytical Plan

Descriptive statistics, including means (with standard deviations) for continuous variables and counts (with proportions) for categorical variables, were reported overall and stratified by disability status.

Ordinal logistic regression examined associations between disability status and past 30-day cannabis and tobacco use, described with adjusted odds ratios (aOR). This approach assumes the odds of being in a higher use category (e.g., daily use) relative to all lower categories are consistent across thresholds. The proportional odds assumption was tested and met. Negative-binomial regression estimated incidence rate ratios (IRRs) of past 30-day binge drinking. Given the right-skewed distribution of binge drinking frequency in the sample, we modeled the outcome as a count variable to estimate associations across the full range of reported frequencies.23 For past 6-month non-medical stimulant and opioid use (any vs. none), logistic regression estimated ORs. Absolute prevalence (%) was reported alongside ORs to show the actual size of the group difference.24 Because the CARMA study used quota-based sampling (50% past-month cannabis users), results were interpreted as within-sample associations rather than national prevalence estimates.

Estimates and 95% confidence intervals (CIs) were reported using a three-model strategy: an unadjusted model, a model adjusted for socio-demographics identified in prior literature, and a fully-adjusted model that included mental health symptoms.25,26 For cannabis outcomes, models also adjusted for state cannabis legalization status.27,28 We conducted sensitivity analyses after excluding participants who endorsed disability without specifying subtype to evaluate whether unavailable subtype information affected associations between disability status and substance use outcomes. In addition, we tested linear trends in substance use outcomes by modeling the number of functional limitations as a continuous term (0,1,2,3,4). Analyses were conducted using SAS v9.4. Statistical tests were two-sided; p-values <0.05 were considered statistically significant.

RESULTS

Sample Characteristics

Of CARMA study participants who completed the baseline Wave 1 survey (n=4,031), 89.0% completed the Wave 3 survey (n=3,588). We excluded participants who did not respond to the disability assessment (n=1) or selected “Prefer not to answer” (n=129), yielding an analytic sample of 3,458 participants. Socio-demographic characteristics of the sample and rates of depression and anxiety symptoms are shown in Table 1. Nearly one-third of the sample self-identified as having a disability (n=1,044, 30.2%). Compared to people without disabilities (PWOD), PWD were more likely to be female, White, have lower levels of educational attainment, be unemployed or employed part-time, reside in rural (vs. urban) areas, and live in states where non-medical cannabis use was not legalized (all ps<0.001).

Table 1.

Socio-demographics by disability status among the study sample (N = 3458).

Disability status
Overall None1
(n = 2414, 69.8%)		 Any
(n = 1044, 39.2%)
N (%) p
Sociodemographic
 Age* 26.4 (4.8) 26.5 (4.8) 26.2 (4.9) 0.09
 Sex at birth
  Female 2103 (61.2) 1420 (59.0) 683 (66.2) <.001
  Male 1334 (38.8) 985 (41.0) 349 (33.8)
 Hispanic
  Non-Hispanic 2762 (81.0) 1927 (81.0) 835 (80.8) 0.91
  Hispanic 650 (19.0) 452 (19.0) 198 (19.2)
 Race
  White 2184 (63.2) 1487 (61.6) 697 (66.8) <.001
  Black 443 (12.8) 300 (12.4) 143 (13.7)
  Asians 489 (14.1) 407 (16.9) 82 (7.9)
  Other2 342 (9.9) 220 (9.1) 122 (11.7)
 Education
  < Bachelor's degree 1884 (54.5) 1161 (48.1) 723 (26.3) <.001
  ≥ Bachelor's degree 1574 (45.5) 1253 (51.9) 321 (30.8)
 Employment status
  Full-time 1272 (36.8) 998 (41.3) 274 (26.3) <.001
  Part-time 451 (13.0) 271 (11.2) 180 (17.2)
  Student 942 (27.2) 671 (27.8) 271 (26.0)
  Unemployed 793 (22.9) 474 (19.6) 319 (30.5)
 Community type^
  Rural 694 (20.2) 442 (18.4) 252 (24.3) <.001
  Suburban/micropolitan 1060 (30.8) 737 (30.7) 323 (31.10
  Urban 1687 (49.0) 1223 (50.9) 464 (44.7)
 Relationship
  Single/other 2034 (58.8) 1351 (56.0) 683 (65.4) <.001
  Married 831 (24.0) 671 (27.8) 160 (15.3)
  Cohabitating 593 (17.2) 392 (16.2) 201 (19.3)
 Parent/has child(ren)
  No 2386 (69.0) 1620 (67.1) 766 (73.4) <.001
  Yes 1072 (31.0) 794 (32.9) 278 (26.6)
 State non-medical cannabis law
  Not legalized 1697 (49.1) 1141 (47.3) 556 (53.3) 0.001
  Legalized3 1761 (50.9) 1273 (52.7) 488 (46.7)
Mental health symptoms
 Depression (assessed by PHQ-8)
  None/Mild (<10) 2240 (64.8) 1805 (74.8) 435 (41.7) <.001
  Moderate (10–14) 617 (17.8) 340 (14.1) 277 (26.5)
  Severe (>15) 601 (17.4) 269 (11.1) 332 (31.8)
 Anxiety (assessed by GAD-7)
  None/Mild (<10) 2313 (66.9) 1833 (75.9) 480 (46.0) <.001
  Moderate (10–14) 640 (18.5) 331 (13.7) 309 (29.6)
  Severe (>15) 505 (14.6) 250 (10.4) 255 (24.4)
Open in a new tab

SD, Standard Deviation. PHQ-8, Patient Health Questionnaire-8; GAD-7, Generalized Anxiety Disorder-7.

*

Mean (SD)

**

Sex at birth: other/prefer not to answer (n = 21). Ethnicity: donť know/refuse to say (n = 46). Community: other (n = 17).

1

Participants who endorsed either the ADA-defined disability item or any of the six functional limitation items (i.e., hearing, vision, cognitive, mobility, or complex activity) were classified as having any disability (vs. none).

2

Race: other (i.e., American Indian or Alaskan Native, Native Hawaiian or Pacific Islander, Multiracial, other).

3

Legalized non-medical adult use as of March 2023: Alaska, Arizona, California, Colorado, Connecticut, District of Columbia, Illinois, Maine, Maryland, Massachusetts, Michigan, Missouri, Montana, Nevada, New Jersey, New Mexico, New York, Oregon, Rhode Island, Vermont, Virginia, Washington.

^

Rural: <10,000; suburban/micropolitan: 10,000–49,999; urban: 50,000+.

P-values were calculated on Pearson CHISQ tests. Bold indicates statistically significant.

The Associations Between Disability Status and Substance Use

In ordinal logistic regression controlling for socio-demographics (e.g., age, gender, race/ethnicity), state-level cannabis legalization, and mental health symptoms, PWD (vs. PWOD) were significantly more likely to report more frequent cannabis use (aOR: 1.44, 95%CI: 1.23–1.69; Table 2). By subtype, more frequent past 30-day cannabis use was significantly associated with mobility/complex activity limitation only (1.80, 1.33–2.44), and ≥2 disabling limitations (1.80, 1.37–2.35), compared with PWOD. By number of limitations, the odds of past 30-day cannabis use were elevated for participants with 1 limitation (1.44, 1.17–1.78), 2 limitations (1.70, 1.28–2.27), and ≥3 limitations (2.41, 1.25–4.64) compared to PWOD.

Table 2.

Multivariable ordinal regression assessing the association between past 30-day marijuana use and disability status (N = 3451*).

Frequency of using cannabis over past 30-day
Overall Non-use
(n = 2052, 59.5%)		 Infrequent
(n = 500, 14.5%)		 Frequent
(n = 544, 15.8%)		 Daily
(n = 355, 10.3%)
N (%) OR (95% CI)1 aOR (95% CI)2 aOR (95% CI)3
Disability status
 None 2411 (69.9) 1532 (74.7) 345 (69.0) 329 (60.5) 205 (57.8) Ref Ref Ref
 Any 1040 (30.1) 520 (25.3) 155 (31.0) 215 (39.5) 150 (42.2) 1.99 (1.69, 2.33) 1.55 (1.29, 1.86) 1.44 (1.23, 1.69)
Disability status, by type4
 None 2411 (69.9) 1532 (82.2) 345 (69.0) 329 (60.5) 205 (66.8) Ref Ref Ref
 Hearing limitation, only 47 (1.5) 26 (1.4) 6 (1.3) 13 (2.6) 2 (0.7) 1.29 (0.73, 2.27) 1.17 (0.65, 2.1) 1.1 (0.61, 1.98)
 Vision limitation, only 36 (1.2) 16 (0.9) 7 (1.6) 9 (1.8) 4 (1.3) 2.06 (1.11, 3.82) 1.57 (0.83, 2.98) 1.52 (0.8, 2.88)
 Cognitive limitation, only 194 (6.2) 101 (5.4) 29 (6.4) 41 (8.3) 23 (7.5) 1.62 (1.23, 2.14) 1.36 (1.02, 1.82) 1.25 (0.93, 1.69)
 Mobility/complex activity limitation, only 179 (5.8) 80 (4.3) 30 (6.7) 40 (8.1) 29 (9.5) 2.16 (1.62, 2.88) 1.91 (1.41, 2.58) 1.80 (1.33, 2.44)
 ≥2 limitations 247 (7.9) 108 (5.8) 34 (7.5) 61 (12.4) 44 (14.3) 2.37 (1.86, 3.03) 2.00 (1.55, 2.58) 1.80 (1.37, 2.35)
Disability status, by number of limitations4
 None 2411 (69.9) 1532 (82.2) 345 (69.0) 329 (60.5) 205 (66.8) Ref Ref Ref
 1 limitation 456 (14.6) 223 (12.0) 72 (16.0) 103 (20.9) 58 (18.9) 1.81 (1.49, 2.19) 1.54 (1.26, 1.89) 1.44 (1.17, 1.78)
 2 limitations 214 (6.9) 97 (5.2) 26 (5.8) 52 (10.6) 39 (12.7) 2.33 (1.79, 3.02) 1.89 (1.44, 2.48) 1.70 (1.28, 2.27)
 ≥ 3 limitations 33 (1.1) 11 (0.6) 8 (1.8) 9 (1.8) 5 (1.6) 2.66 (1.41, 5.01) 2.70 (1.41, 5.17) 2.41 (1.25, 4.64)
Open in a new tab

(a)OR, (adjusted) Odds Ratio; CI, Confidence Interval; Ref, Reference.

*

Participants who did not report the use status of cannabis (i.e., cannabis) were excluded from the analysis (n = 7).

1

The effect sizes of estimates were obtained from unadjusted ordinal logistic regression.

2

The model was adjusted for state legalization of cannabis, age, sex assigned at birth, ethnicity, race, educational attainment, employment status, community type (e.g., rural, urban), relationship status, and whether participants had children.

3

The model was additionally adjusted for depression and anxiety symptoms.

4

A total of 338 participants who did not provide data on any specific functional limitation were excluded from the model.

Bold indicates statistically significant.

PWD (vs. PWOD) were significantly more likely to report more days of tobacco use in the past 30-days (aOR: 1.61, 95%CI: 1.33–1.93; Table 3) in the fully-adjusted model. All disability subtypes, except cognitive, were positively associated with more frequent tobacco use, compared to PWOD, with the highest odds in individuals with vision limitation only (2.42, 1.28–4.58). Individuals with 1 limitation (1.56, 1.26–1.92) and 2 limitations (1.69, 1.26–2.27) were more likely to report a higher frequency of tobacco use compared to PWOD; however, having ≥3 limitations was not significantly associated with the outcome (1.80, 0.92–3.55).

Table 3.

Multivariable ordinal regression assessing the association between past 30-day tobacco users and disability status (N = 3458).

Days of using tobacco over past 30-day
Overall Non-use
(n = 2150, 62.2%)		 1–9 days
(n = 500, 14.5%)		 10–24 days
(n = 215, 6.2%)		 ≥25 days
(n = 593, 17.2%)
N (%) OR (95% CI)1 aOR (95% CI)2 aOR (95% CI)3
Disability status
 None 2414 (69.8) 1598 (74.3) 330 (66.0) 132 (61.4) 354 (59.7) Ref Ref Ref
 Any 1044 (30.2) 552 (25.7) 170 (34.0) 83 (38.6) 239 (40.3) 2.20 (1.87, 2.59) 1.57 (1.34, 1.83) 1.61 (1.33, 1.93)
Disability status, by type4
 None 2414 (77.4) 1598 (82.8) 330 (66.0) 132 (61.4) 354 (59.7) Ref Ref Ref
 Hearing limitation, only 47 (1.5) 21 (1.1) 10 (2.2) 8 (4.1) 8 (1.5) 2.09 (1.21, 3.60) 1.89 (1.08, 3.32) 1.76 (1.01, 3.10)
 Vision limitation, only 36 (1.2) 11 (0.6) 12 (2.7) 4 (2.0) 9 (1.7) 3.08 (1.67, 5.67) 2.65 (1.40, 5.01) 2.42 (1.28, 4.58)
 Cognitive limitation, only 195 (6.3) 111 (5.8) 21 (4.6) 16 (8.2) 47 (8.7) 1.62 (1.22, 2.14) 1.38 (1.03, 1.86) 1.13 (0.84, 1.54)
 Mobility/complex activity limitation, only 181 (5.8) 87 (4.5) 28 (6.2) 14 (7.1) 52 (9.6) 2.36 (1.77, 3.14) 2.11 (1.56, 2.86) 1.86 (1.37, 2.53)
 ≥2 limitations 247 (7.9) 103 (5.3) 52 (11.5) 22 (11.2) 70 (13.0) 2.52 (1.97, 3.23) 2.14 (1.65, 2.77) 1.69 (1.28, 2.22)
Disability status, by number of limitations4
 None 2414 (77.4) 1598 (82.8) 330 (66.0) 132 (61.4) 354 (59.7) Ref Ref Ref
 1 limitation 459 (14.7) 230 (11.9) 71 (15.6) 42 (21.4) 116 (21.5) 2.04 (1.69, 2.48) 1.79 (1.46, 2.19) 1.56 (1.26, 1.92)
 2 limitations 214 (6.9) 88 (4.6) 45 (9.9) 19 (9.7) 62 (11.5) 2.62 (2.01, 3.41) 2.13 (1.61, 2.81) 1.69 (1.26, 2.27)
 ≥ 3 limitations 33 (1.1) 15 (0.8) 7 (1.6) 3 (1.5) 8 (1.5) 1.98 (1.03, 3.81) 2.20 (1.13, 4.31) 1.80 (0.92, 3.55)
Open in a new tab

(a)OR, (adjusted) Odds Ratio; CI, Confidence Interval; Ref, Reference.

1

The effect sizes of estimates were obtained from unadjusted ordinal logistic regression.

2

The model was adjusted for socio-demographics, including age, sex assigned at birth, ethnicity, race, educational attainment, employment status, community type (e.g., rural, urban), relationship status, and whether participants had children.

3

The model was additionally adjusted for depression and anxiety symptoms.

4

A total of 338 participants who did not provide data on any specific functional limitation were excluded from the model.

Bold indicates statistically significant.

The results from unadjusted regression model indicated that having any disability was positively associated with binge drinking; however, the association was non-significant in the adjusted models (IRR: 0.94, 95%CI: 0.77–1.14; Table 4). Compared to PWOD, those with mobility/complex activity limitation only had a higher rate of binge drinking in the past 30-days (1.55, 1.08–2.21), while those with cognitive limitation only had a lower rate (0.67, 0.47–0.97). By number of disabling limitations, no significant association was found with binge drinking frequency after controlling for socio-demographics and mental health symptoms.

Table 4.

Multivariable negative-binominal regression assessing the association between past 30-day binge drinking and disability status (N = 3458).

Days of binge drinking alcohol over past 30-day
Mean (SD) IRR (95% CI)1 IRR (95% CI)2 IRR (95% CI)3
Disability status
 None 1.27 (3.12) Ref Ref Ref
 Any 1.52 (3.51) 1.20 (1.01, 1.43) 1.19 (0.99, 1.43) 0.94 (0.77, 1.14)
Disability status, by type4
 None 1.27 (3.12) Ref Ref Ref
 Hearing limitation, only 2.36 (3.70) 1.86 (0.96, 3.62) 1.78 (0.91, 3.45) 1.54 (0.79, 2.97)
 Vision limitation, only 2.19 (3.01) 1.73 (0.81, 3.70) 1.60 (0.75, 3.43) 1.45 (0.68, 3.10)
 Cognitive limitation, only 1.11 (2.52) 0.87 (0.62, 1.24) 0.89 (0.63, 1.26) 0.67 (0.47, 0.97)
 Mobility/complex activity limitation, only 2.10 (4.33) 1.66 (1.17, 2.36) 1.85 (1.29, 2.64) 1.55 (1.08, 2.21)
 ≥2 limitations 2.02 (4.01) 1.60 (1.18, 2.16) 1.51 (1.11, 2.05) 1.14 (0.83, 1.57)
Disability status, by number of limitations4
 None 1.27 (3.12) Ref Ref Ref
 1 limitation 1.71 (3.52) 1.35 (1.07, 1.71) 1.40 (1.11, 1.78) 1.16 (0.91, 1.49)
 2 limitations 1.95 (4.15) 1.54 (1.11, 2.13) 1.43 (1.03, 1.98) 1.09 (0.77, 1.53)
 ≥ 3 limitations 2.48 (2.97) 1.96 (0.89, 4.32) 2.00 (0.91, 4.41) 1.59 (0.72, 3.48)
Open in a new tab

SD, Standard Deviation; IRR, Incidence Rate Ratio; CI, Confidence Interval; Ref, Reference.

1

The effect sizes of estimates were obtained from unadjusted negative-binominal regression.

2

The model was adjusted for socio-demographics, including age, sex assigned at birth, ethnicity, race, educational attainment, employment status, community type (e.g., rural, urban), relationship status, and whether participants had children.

3

The model was additionally adjusted for depression and anxiety symptoms.

4

A total of 338 participants who did not provide data on any specific functional limitation were excluded from the model.

Bold indicates statistically significant.

Binary logistic regression results indicated PWD (vs. PWOD) were significantly more likely to report non-medical stimulant use in the past 6 months (aOR: 3.02, 95%CI: 2.23–4.09; Table 5) in the fully-adjusted model. All disability subtypes, except cognitive, were positively associated with stimulant use; the highest odds were observed among individuals with ≥2 limitation types (4.65, 3.13–6.90) and those with vision limitation only (4.55, 1.89–10.93). Compared to PWOD, individuals with 1 limitation (2.29, 1.61–3.27), 2 limitations (4.49, 2.96–6.79), and ≥3 limitations (6.83, 2.80–16.64) were more likely to report past 6-month non-medical stimulant use.

Table 5.

Multivariable logistic regression assessing the association between past 6-month non-medical stimulant use and disability status (N = 3209*).

Non-medical stimulants use over past 6-month
Overall No
(n = 2930, 91.3%)		 Yes
(n = 279, 8.7%)
N (%) Absolute % OR (95% CI)1 aOR (95% CI)2 aOR (95% CI)3
Disability status
 None 2325 (72.5) 2202 (75.2) 123 (44.1) 5.3 Ref Ref Ref
 Any 884 (27.6) 728 (24.8) 156 (55.9) 17.6 3.84 (2.99, 4.93) 4.52 (3.41, 6.00) 3.02 (2.23, 4.09)
Disability status, by type4
 None 2325 (79.4) 2202 (75.2) 123 (44.1) 5.3 Ref Ref Ref
 Hearing limitation, only 40 (1.4) 32 (1.2) 8 (3.2) 20.0 4.48 (2.02, 9.93) 4.29 (1.91, 9.65) 3.47 (1.51, 7.99)
 Vision limitation, only 31 (1.1) 22 (0.8) 9 (3.6) 29.0 7.33 (3.30, 16.25) 5.46 (2.33, 12.76) 4.55 (1.89, 10.93)
 Cognitive limitation, only 163 (5.6) 144 (5.4) 19 (7.7) 11.7 2.36 (1.42, 3.94) 2.27 (1.35, 3.83) 1.41 (0.82, 2.43)
 Mobility/complex activity limitation, only 163 (5.6) 138 (5.2) 25 (10.1) 15.3 3.24 (2.04, 5.15) 3.45 (2.14, 5.57) 2.71 (1.67, 4.41)
 ≥2 limitations 207 (7.1) 144 (5.4) 63 (25.5) 30.4 7.83 (5.53, 11.09) 7.84 (5.43, 11.33) 4.65 (3.13, 6.90)
Disability status, by number of limitations4
 None 2325 (79.4) 2202 (75.2) 123 (44.1) 5.3 Ref Ref Ref
 1 limitation 397 (13.6) 336 (12.5) 61 (24.7) 15.4 3.25 (2.34, 4.51) 3.17 (2.25, 4.45) 2.29 (1.61, 3.27)
 2 limitations 182 (6.2) 128 (4.8) 54 (21.9) 29.7 7.55 (5.24, 10.89) 7.51 (5.10, 11.06) 4.49 (2.96, 6.79)
 ≥ 3 limitations 25 (0.9) 16 (0.6) 9 (3.6) 36.0 10.07 (4.36, 23.26) 10.48 (4.40, 24.92) 6.83 (2.80, 16.64)
Open in a new tab

(a)OR, (adjusted) Odds Ratio; CI, Confidence Interval; Ref, Reference.

*

Participants who used stimulants that were all prescribed to them were excluded from the analysis (n = 249).

1

The effect sizes of estimates were obtained from unadjusted logistic regression.

2

The model was adjusted for socio-demographics, including age, sex assigned at birth, ethnicity, race, educational attainment, employment status, community type (e.g., rural, urban), relationship status, and whether participants had children.

3

The model was additionally adjusted for depression and anxiety symptoms.

4

A total of 280 participants who did not provide data on specific functional limitation were excluded from the model.

Bold indicates statistically significant.

PWD (vs. PWOD) had significantly higher odds of reporting past 6-month non-medical opioid use (aOR: 3.00, 95%CI: 2.11–4.28; Table 6) in the fully-adjusted model. By disability subtype, those with vision limitation (6.54, 2.54–16.85), mobility/complex activity limitation (4.07, 2.31–7.17), and ≥2 limitation types (5.04, 3.10–8.22) were significantly more likely to report non-medical opioid use. Compared to PWOD, individuals with 1 limitation (2.65; 1.70–4.13), 2 limitations (4.96, 2.97–8.29), and ≥3 limitations (6.95, 2.49–19.36) were more likely to report past 6-month non-medical opioid use.

Table 6.

Multivariable logistic regression assessing the association between past 6-month non-medical opioids use and disability status (N = 3282*).

Non-medical opioids use over past 6-month
Overall No
(n = 3109, 94.7%)		 Yes
(n = 173, 5.3%)
N (%) Absolute % OR (95% CI)1 aOR (95% CI)2 aOR (95% CI)3
Disability status
 None 2333 (71.1) 2264 (72.8) 69 (39.9) 3.0 Ref Ref
 Any 949 (28.9) 845 (27.2) 104 (60.1) 11.0 4.62 (3.23, 6.60) 4.20 (3.00, 5.88) 3.00 (2.11, 4.28)
Disability status, by type3
 None 2333 (71.1) 2264 (72.8) 69 (39.9) 3.0 Ref Ref Ref
 Hearing limitation, only 45 (1.5) 43 (1.5) 2 (1.3) 4.4 1.54 (0.36, 6.48) 1.6 (0.37, 6.83) 1.36 (0.31, 5.89)
 Vision limitation, only 34 (1.1) 26 (0.9) 8 (5.2) 23.5 8.69 (3.65, 20.7) 7.87 (3.12, 19.85) 6.54 (2.54, 16.85)
 Cognitive limitation, only 177 (6.0) 166 (5.9) 11 (7.2) 6.2 2.14 (1.11, 4.12) 2.12 (1.07, 4.20) 1.35 (0.67, 2.75)
 Mobility/complex activity limitation, only 164 (5.5) 144 (5.1) 20 (13.1) 12.2 4.74 (2.80, 8.04) 5.10 (2.92, 8.90) 4.07 (2.31, 7.17)
 ≥2 limitations 219 (7.4) 176 (6.2) 43 (28.1) 19.6 7.69 (5.07, 11.66) 8.40 (5.36, 13.18) 5.04 (3.10, 8.22)
Disability status, by number of limitations3
 None 2333 (71.1) 2264 (72.8) 69 (39.9) 3.0 Ref Ref Ref
 1 limitation 387 (13.0) 350 (12.4) 37 (24.2) 9.6 3.49 (2.33, 5.23) 3.56 (2.32, 5.46) 2.65 (1.70, 4.13)
 2 limitations 191 (6.4) 155 (5.5) 36 (23.5) 18.8 7.48 (4.82, 11.6) 8.16 (5.07, 13.13) 4.96 (2.97, 8.29)
 ≥ 3 limitations 28 (0.9) 21 (0.7) 7 (4.6) 25.0 9.22 (3.61, 23.56) 10.06 (3.71, 27.25) 6.95 (2.49, 19.36)
Open in a new tab

(a)OR, (adjusted) Odds Ratio; CI, Confidence Interval; Ref, Reference.

*

Participants who used opioids that were all prescribed to them were excluded from the analysis (n = 176).

1

The effect sizes of estimates were obtained from unadjusted logistic regression.

2

The model was adjusted for socio-demographics, including age, sex assigned at birth, ethnicity, race, educational attainment, employment status, community type (e.g., rural, urban), relationship status, and whether participants had children.

3

The model was additionally adjusted for depression and anxiety symptoms.

4

A total of 310 participants who did not provide data on specific functional limitation were excluded from the model.

Bold indicates statistically significant.

After excluding participants without disability subtype information for the sensitivity analysis, disability status remained significantly associated with substance use in the fully-adjusted model, except for binge drinking (Supplemental Table 1). Significant linear trends were also observed between the number of disability limitations and substance use outcomes, except for binge drinking (Supplemental Table 2).

DISCUSSION

Our study offers a novel contribution to the literature on substance use in PWD via a national cohort of U.S. young adults. Extending prior work,11 we offer a granular analysis of use patterns by mapping recent use of five substance classes against disability subtypes and number of limitations. Among our sample, having any disability was associated with significantly higher odds of past 30-day cannabis and tobacco use and past 6-month non-medical stimulant and opioid use. Likelihood of use also increased with the number of disabling limitations. Binge drinking however did not follow similar patterns. These findings suggest that disability status is generally correlated with higher rates and frequency of substance use, though patterns vary by substance and disability characteristics.

PWD, compared to PWOD, reported more frequent past 30-day cannabis use, consistent with NSDUH findings. The association was strongest among those with mobility/complex activity limitations, consistent with findings that individuals with physical disabilities often use cannabis for pain and functional relief.29 Further, having multiple disabling limitations signaled more complex health needs, consistent with self-medication patterns,30 though the reverse remains possible, where cannabis can exacerbate disability severity. While cannabis motivations were not explored, self-medicating has been strongly linked to higher cannabis use frequencies and more use-related problems.31 Clinicians must screen for cannabis use in PWD, assess whether cannabis is being used to cope with pain or disability, and discern if limited access to healthcare is contributing to higher frequency of cannabis use.

All PWD subgroups had higher odds of tobacco use than PWOD, even after adjusting for covariates, suggesting disability is independently associated with smoking tobacco. Similar patterns appeared in NSDUH32 and the National Health Interview Survey,33 reinforcing the external validity of our findings. A stepwise pattern in tobacco use frequency emerged with 0, 1, and 2 disabling limitations, potentially consistent with patterns of stress or discrimination linked to increased smoking,34 or conversely with smoking complications contributing to disability. However, use decreased among those with ≥3 limitations, potentially reflecting limited access to tobacco at higher levels of impairment, increased exposure to medical settings with cessation resources, or increased motivation to quit given extensive functional impact.35,36

Patterns of binge drinking differed from the other substances studied. Although participants with any disability reported more binge-drinking days in unadjusted analyses, this association was attenuated and no longer statistically significant after adjustment for sociodemographic characteristics and, particularly, depression and anxiety symptoms. However, mobility or complex activity limitation remained associated with a higher rate of binge-drinking days in the fully-adjusted model. Heavy episodic drinking is particularly concerning in this group, as it compounds fall and injury risk,37 and can interact with medications commonly used to manage pain.38 Conversely, participants in our sample with cognitive limitations drank less often than PWOD, which may reflect less independent access to alcohol or more structured daily environments.39,40 With risk patterns diverging based on disability subtype, binge drinking exhibits why such patterns are worth mapping by both substance and limitation.

Disability was strongly associated with non-medical stimulant use: PWD had three times the odds of past 6-month use compared to PWOD, and those with ≥2 disability subtypes had almost five times higher odds of such stimulant use. This stark magnitude is explained at least partially by our sample, comprised of only young adults and stratified for recent cannabis use, capturing higher-risk groups of PWD.41–45 The higher odds of recent stimulant use with 1 and 2 disabling limitations may reflect attempts to cope with increasing functional demands, such as fatigue or cognitive effort, although stimulant exposure itself can negatively impact function over time,46 representing considerations that should factor into clinicians’ counseling practices. The magnitude of these associations should be interpreted cautiously, given that large aORs and wide confidence intervals reflect small subgroup sizes, especially in participants with two or more disability subtypes.47–49

Regarding non-medical opioid use, PWD had three times the odds of past 6-month use compared to PWOD, again consistent with NSDUH findings.10,19,50 Participants with mobility/complex activity limitations had almost three times higher odds of non-medical opioid use, while those with visual limitations had over four times higher odds, compared to PWOD. Such gradients indicate that opioid use risk is not tied solely to pain and may co-occur with sensory or functional challenges, though the direction of this relationship is unclear in cross-sectional data. As such, screening for opioid use alone is insufficient unless paired with assessments of functional barriers that shape how people manage pain and disability.

Our findings underscore the need for substance use screening and prevention efforts that consider disability-related needs. While screening mechanisms for substance use in the general population account for multiple behavioral and clinical factors, our findings indicate that PWD face distinct sensory, mobility, and accessibility challenges that must be incorporated into routine assessment and clinical decision-making. It is important to determine whether substance use is occurring to cope with features of their disability. Conversations about self-medication should also be framed such that PWD feel safe disclosing why they use a substance.

At a system level, accessible screening instruments specifically adapted for PWD should be utilized. For example, the Drug Abuse Screening Test51 can be administered verbally for those with motor limitations, and there is a Substance Abuse Screener validated in American Sign Language (SAS-ASL).52 Providers should be trained in disability-informed topics to improve cultural competence, and substance use treatment environments should be made accessible to PWD, encouraging follow-through in substance use recovery. Developing preventive educational materials tailored to the needs of PWD in accessible formats can educate PWD on healthy coping mechanisms and substance use prevention. Such materials should be developed with consultation from PWD themselves, to ensure they are culturally sensitive and relevant to the community.

Further studies should attempt to examine temporal relationships between substance use and disability onset. Gathering additional information regarding disability severity, use motivations, and co-use patterns of multiple substances could provide additional insight into the relationships between disability and substance use. Subjective assessments could be integrated into substance use screening tools to better detect self-medication related to medical undertreatment.

Limitations

The CARMA study sample was balanced using quotas for past-month cannabis use, thus oversampling young adults who use cannabis, and likely co-selecting for higher use of other substances. As a result, prevalence and effect estimates from this sample are not generalizable to the U.S. young adult population and the magnitude of reported associations likely exceeds those in a probability sample. Additionally, quota-based sampling can inflate effect size estimates, meaning that large associations (e.g., for stimulant and opioid use) should be interpreted with caution. However, estimating prevalence was not the goal of the study, as NSDUH provides such data; we hoped to instead understand the relationship between disability and substance use.

Next, our cross-sectional design precludes causal inference. Bidirectionality is plausible across the substances examined: for example, cannabis or opioids may be used to cope with pain or functional limitations, though sustained use can also produce or worsen cognitive, motor, or physiologic impairment; and shared upstream factors may drive both disability status and substance use independently.53–55 We lacked data on diagnosed mental health comorbidities, chronic pain, and healthcare access (insurance, utilization), all of which are plausibly associated with both disability status and substance use and potentially contain residual confounding effects.

We identified 338 participants who reported a disability without specifying a subtype, limiting precision in our disability classification scheme. Accessibility barriers may have prevented some individuals with disabilities from completing the survey. All disability and substance use data were self-reported, which can introduce recall and social desirability bias. Substance use behaviors may be also underreported due to stigma, though self-reported data have been validated for both disability status56–58 and substance use.59–62

Conclusion

PWD reported more frequent past 30-day cannabis and tobacco use, and more non-medical stimulant and opioid use in the past 6 months, compared to PWOD. Because use patterns differed by substance and by type and number of disabling limitations, clinician screening and prevention practices should account for the specific limitations involved, both to identify those at risk for substance use or overdose, and to understand what drives their substance use. When services such as rehabilitation therapies, assistive devices, or behavioral healthcare are inaccessible, substance use can become one of the few available ways to cope. Meaningful progress in substance use prevention and treatment will require confronting the structural and functional barriers that limit care for PWD.

Supplementary Material

Supplementary Tables 1-2

Highlights.

  • Young adults with disabilities had 1.4 to 3.0x higher odds of recent cannabis, tobacco, and non-medical stimulant or opioid use vs. those without disabilities

  • Patterns of substance use varied greatly by disability subtype and number of limitations

  • Substance use screening for young adults with disabilities should include assessment of pain, sensory, and functional barriers to care

References

  • 1.Hall WD, Patton G, Stockings E, et al. Why young people’s substance use matters for global health. Lancet Psychiatry. 2016;3(3):265–279. doi: 10.1016/S2215-0366(16)00013-4 [DOI] [PubMed] [Google Scholar]
  • 2.Kacha-Ochana A, Jones CM, Green JL, et al. Characteristics of Adults Aged ≥18 Years Evaluated for Substance Use and Treatment Planning - United States, 2019. MMWR Morb Mortal Wkly Rep. 2022;71(23):749–756. doi: 10.15585/mmwr.mm7123a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.McCabe SE, Arterberry BJ, Dickinson K, et al. Assessment of Changes in Alcohol and Marijuana Abstinence, Co-Use, and Use Disorders Among US Young Adults From 2002 to 2018. JAMA Pediatr. 2021;175(1):64–72. doi: 10.1001/jamapediatrics.2020.3352 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Twohey M, Ivory D, Kessler C. As America’s Marijuana Use Grows, So Do the Harms. N Y Times. October 4, 2024. Accessed November 12, 2025. https://www.nytimes.com/2024/10/04/us/cannabis-marijuana-risks-addiction.html [Google Scholar]
  • 5.Compton WM, Flannagan KSJ, Silveira ML, et al. Tobacco, Alcohol, Cannabis, and Other Drug Use in the US Before and During the Early Phase of the COVID-19 Pandemic. JAMA Netw Open. 2023;6(1):e2254566. doi: 10.1001/jamanetworkopen.2022.54566 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Hoots B, Vivolo-Kantor A, Seth P. The rise in non-fatal and fatal overdoses involving stimulants with and without opioids in the United States. Addict Abingdon Engl. 2020;115(5):946–958. doi: 10.1111/add.14878 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Nelson SE, Van Ryzin MJ, Dishion TJ. Alcohol, marijuana, and tobacco use trajectories from age 12 to 24 years: demographic correlates and young adult substance use problems. Dev Psychopathol. 2015;27(1):253–277. doi: 10.1017/S0954579414000650 [DOI] [PubMed] [Google Scholar]
  • 8.Stone AL, Becker LG, Huber AM, Catalano RF. Review of risk and protective factors of substance use and problem use in emerging adulthood. Addict Behav. 2012;37(7):747–775. doi: 10.1016/j.addbeh.2012.02.014 [DOI] [PubMed] [Google Scholar]
  • 9.SAMHSA. Key Substance Use and Mental Health Indicators in the United States: Results from the 2024 National Survey on Drug Use and Health. Substance Abuse and Mental Health Services Administration; 2025. [Google Scholar]
  • 10.Casseus M, West B, Graber JM, Wackowski O, Cooney JM, Lee HS. Disparities in illicit drug use and disability status among a nationally representative sample of U.S. college students. Disabil Health J. 2021;14(1):100949. doi: 10.1016/j.dhjo.2020.100949 [DOI] [PubMed] [Google Scholar]
  • 11.Schulz JA, Gimm G, West JC, Kock L, Villanti AC. Patterns of Alcohol, Marijuana, and Tobacco Use Among U.S. Adolescents and Young Adults by Disability Status: 2015–2019. J Adolesc Health Off Publ Soc Adolesc Med. 2024;74(4):764–773. doi: 10.1016/j.jadohealth.2023.09.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Roux AM, Tao S, Marcus S, Lushin V, Shea LL. A national profile of substance use disorder among Medicaid enrollees on the autism spectrum or with intellectual disability. Disabil Health J. 2022;15(2, Supplement):101289. doi: 10.1016/j.dhjo.2022.101289 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Berg CJ, Romm KF, LoParco CR, et al. Young Adults’ Experiences with Cannabis Retailer Marketing and Related Practices: Differences Among Sociodemographic Groups and Associations with Cannabis Use-related Outcomes. J Racial Ethn Health Disparities. Published online July 15, 2024. doi: 10.1007/s40615-024-02092-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.US Census Bureau. How Disability Data are Collected from The American Community Survey. Census.gov. November 21, 2021. Accessed August 7, 2025. https://www.census.gov/topics/health/disability/guidance/data-collection-acs.html [Google Scholar]
  • 15.Jeffers AM, Glantz S, Byers A, Keyhani S. Sociodemographic Characteristics Associated With and Prevalence and Frequency of Cannabis Use Among Adults in the US. JAMA Netw Open. 2021;4(11):e2136571. doi: 10.1001/jamanetworkopen.2021.36571 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Sánchez-Romero LM, Cadham CJ, Hirschtick JL, et al. A comparison of tobacco product prevalence by different frequency of use thresholds across three US surveys. BMC Public Health. 2021;21(1):1203. doi: 10.1186/s12889-021-11283-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Center for Behavioral Health Statistics and Quality. 2019 NSDUH Methodological Summary and Definitions. Substance Abuse and Mental Health Services Administration; 2020. Accessed August 18, 2025. https://www.samhsa.gov/data/report/2019-methodological-summary-and-definitions [Google Scholar]
  • 18.Gimm G, Parekh T, Kitsantas P. Assessing the prevalence and risk factors of marijuana use in adults with disabilities. Addict Behav. 2023;138:107559. doi: 10.1016/j.addbeh.2022.107559 [DOI] [PubMed] [Google Scholar]
  • 19.Hong YR, Xie Z, Yadav S, Tanner R, Striley C, Marlow NM. Opioid Use Behaviors Among People With Disability in the United States: An Analysis of the National Survey on Drug Use and Health. J Addict Med. 2023;17(1):e27. doi: 10.1097/ADM.0000000000001031 [DOI] [PubMed] [Google Scholar]
  • 20.Xie Z, Stallings-Smith S, Stetten N, Hamadi HY, Marlow NM. Marijuana use disorder among adults with functional disabilities—A US population-based cross-sectional study. Am J Addict. 2024;33(1):26–35. doi: 10.1111/ajad.13485 [DOI] [PubMed] [Google Scholar]
  • 21.Kroenke K, Strine TW, Spitzer RL, Williams JBW, Berry JT, Mokdad AH. The PHQ-8 as a measure of current depression in the general population. J Affect Disord. 2009;114(1–3):163–173. doi: 10.1016/j.jad.2008.06.026 [DOI] [PubMed] [Google Scholar]
  • 22.Rutter LA, Brown TA. Psychometric Properties of the Generalized Anxiety Disorder Scale-7 (GAD-7) in Outpatients with Anxiety and Mood Disorders. J Psychopathol Behav Assess. 2017;39(1):140–146. doi: 10.1007/s10862-016-9571-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Altman DG, Royston P. The cost of dichotomising continuous variables. BMJ. 2006;332(7549):1080. doi: 10.1136/bmj.332.7549.1080 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Zavala S, Stout JE. Understanding and Communicating Risk: Assessing Both Relative and Absolute Risk Is Absolutely Necessary. JID Innov. 2022;2(2):100097. doi: 10.1016/j.xjidi.2022.100097 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Grant BF, Saha TD, Ruan WJ, et al. Epidemiology of DSM-5 Drug Use Disorder: Results From the National Epidemiologic Survey on Alcohol and Related Conditions–III. JAMA Psychiatry. 2016;73(1):39–47. doi: 10.1001/jamapsychiatry.2015.2132 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Kelly LM, Flori JN, Sabella KA, Zajac K. Demographic Differences Among Adults and Emerging Adults With, Without, and in Recovery from Alcohol and Other Drug Use Disorders. Subst Use Misuse. 2026;61(6):818–828. doi: 10.1080/10826084.2025.2577281 [DOI] [PubMed] [Google Scholar]
  • 27.Cerdá M, Mauro C, Hamilton A, et al. Association Between Recreational Marijuana Legalization in the United States and Changes in Marijuana Use and Cannabis Use Disorder From 2008 to 2016. JAMA Psychiatry. 2020;77(2):165–171. doi: 10.1001/jamapsychiatry.2019.3254 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Berg CJ, LoParco CR, Romm KF, et al. Cannabis use characteristics and associations with problematic use outcomes, quitting-related factors, and mental health among US young adults. Subst Abuse Treat Prev Policy. 2025;20(1):1. doi: 10.1186/s13011-025-00634-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Ware MA, Wang T, Shapiro S, et al. Cannabis for the Management of Pain: Assessment of Safety Study (COMPASS). J Pain. 2015;16(12):1233–1242. doi: 10.1016/j.jpain.2015.07.014 [DOI] [PubMed] [Google Scholar]
  • 30.Yang KH, Tam RM, Satybaldiyeva N, et al. Trends in past-month cannabis use among US adults across a range of disabilities and health conditions, 2015–2019. Prev Med. 2023;177:107768. doi: 10.1016/j.ypmed.2023.107768 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Bresin K, Mekawi Y. Do marijuana use motives matter? Meta-analytic associations with marijuana use frequency and problems. Addict Behav. 2019;99:106102. doi: 10.1016/j.addbeh.2019.106102 [DOI] [PubMed] [Google Scholar]
  • 32.Casseus M, Graber JM, West B, Wackowski O. Tobacco use disparities and disability among US college students. J Am Coll Health. 2022;70(7):2079–2084. doi: 10.1080/07448481.2020.1842425 [DOI] [PubMed] [Google Scholar]
  • 33.Schulz JA, West JC, Hall JP, Villanti AC. Disparities in Tobacco Use by Disability and Type: Findings From the 2019 National Health Interview Survey. Am J Prev Med. 2022;63(4):552–563. doi: 10.1016/j.amepre.2022.05.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Slopen N, Kontos EZ, Ryff CD, Ayanian JZ, Albert MA, Williams DR. Psychosocial stress and cigarette smoking persistence, cessation, and relapse over 9–10 years: a prospective study of middle-aged adults in the United States. Cancer Causes Control CCC. 2013;24(10):1849–1863. doi: 10.1007/s10552-013-0262-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Kalkhoran S, Kruse GR, Chang Y, Rigotti NA. Smoking-Cessation Efforts by US Adult Smokers with Medical Comorbidities. Am J Med. 2018;131(3):318.e1–318.e8. doi: 10.1016/j.amjmed.2017.09.025 [DOI] [PubMed] [Google Scholar]
  • 36.Rojewski AM, Baldassarri S, Cooperman NA, et al. Exploring Issues of Comorbid Conditions in People Who Smoke. Nicotine Tob Res Off J Soc Res Nicotine Tob. 2016;18(8):1684–1696. doi: 10.1093/ntr/ntw016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Aram J, Slopen N, Cosgrove C, Arria A, Liu H, Dallal CM. Self-Reported Disability Type and Risk of Alcohol-Induced Death - A Longitudinal Study Using Nationally Representative Data. Subst Use Misuse. 2024;59(9):1323–1330. doi: 10.1080/10826084.2024.2340993 [DOI] [PubMed] [Google Scholar]
  • 38.Witkiewitz K, Vowles KE. Alcohol and Opioid Use, Co-Use, and Chronic Pain in the Context of the Opioid Epidemic: A Critical Review. Alcohol Clin Exp Res. 2018;42(3):478–488. doi: 10.1111/acer.13594 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Barrett N, Paschos D. Alcohol-related problems in adolescents and adults with intellectual disabilities. Curr Opin Psychiatry. 2006;19(5):481. doi: 10.1097/01.yco.0000238474.07078.41 [DOI] [PubMed] [Google Scholar]
  • 40.Fulda KG, Walters ST, Barnett TE, et al. Substance use disorder among people with intellectual and developmental disabilities: a narrative review. Am J Drug Alcohol Abuse. 2026;52(1):18–32. doi: 10.1080/00952990.2025.2603241 [DOI] [PubMed] [Google Scholar]
  • 41.Abadie R, Cano M. Health disparities and injection drug use behaviors among adults with and without disabilities in the National Survey on drug use and health, 2015–2019. Disabil Health J. 2025;18(1):101715. doi: 10.1016/j.dhjo.2024.101715 [DOI] [PubMed] [Google Scholar]
  • 42.Garcia JM, Ross-Cypcar SM, Haegele JA. Differences in Prescription Drug Misuse Among U.S. Adults With and Without Disabilities. Am J Prev Med. 2025;68(5):954–963. doi: 10.1016/j.amepre.2025.01.025 [DOI] [PubMed] [Google Scholar]
  • 43.Fergusson DM, Boden JM, Horwood LJ. Cannabis use and other illicit drug use: testing the cannabis gateway hypothesis. Addiction. 2006;101(4):556–569. doi: 10.1111/j.1360-0443.2005.01322.x [DOI] [PubMed] [Google Scholar]
  • 44.Wilson J, Mills K, Freeman TP, Sunderland M, Visontay R, Marel C. Weeding out the truth: a systematic review and meta-analysis on the transition from cannabis use to opioid use and opioid use disorders, abuse or dependence. Addiction. 2022;117(2):284–298. doi: 10.1111/add.15581 [DOI] [PubMed] [Google Scholar]
  • 45.Degenhardt L, Hall W, Lynskey M. The relationship between cannabis use and other substance use in the general population. Drug Alcohol Depend. 2001;64(3):319–327. doi: 10.1016/S0376-8716(01)00130-2 [DOI] [PubMed] [Google Scholar]
  • 46.Lappin JM, Sara GE. Psychostimulant use and the brain. Addiction. 2019;114(11):2065–2077. doi: 10.1111/add.14708 [DOI] [PubMed] [Google Scholar]
  • 47.Czeisler MÉ. Mental Health and Substance Use Among Adults with Disabilities During the COVID-19 Pandemic — United States, February–March 2021. MMWR Morb Mortal Wkly Rep. 2021;70. doi: 10.15585/mmwr.mm7034a3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Aitken Z, Simpson JA, Gurrin L, Bentley R, Kavanagh AM. Do material, psychosocial and behavioural factors mediate the relationship between disability acquisition and mental health? A sequential causal mediation analysis. Int J Epidemiol. 2018;47(3):829–840. doi: 10.1093/ije/dyx277 [DOI] [PubMed] [Google Scholar]
  • 49.Wetmore JB, Chernov C, Perlman SE, Borrell LN. Associations of Health Conditions and Health-Related Determinants with Disability among New York City Adult Residents. Ethn Dis. 2021;31(3):445–452. doi: 10.18865/ed.31.3.445 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Lauer EA, Henly M, Brucker DL. Prescription opioid behaviors among adults with and without disabilities – United States, 2015–2016. Disabil Health J. 2019;12(3):519–522. doi: 10.1016/j.dhjo.2018.12.001 [DOI] [PubMed] [Google Scholar]
  • 51.Centre for Addiction and Mental Health. Drug Abuse Screening Test. Centre for Addiction and Mental Health; 1982. [Google Scholar]
  • 52.Guthmann D, Lazowski LE, Moore D, Heinemann AW, Embree J. Validation of the substance abuse screener in American Sign Language (SAS--ASL). Rehabil Psychol. 2012;57(2):140–148. doi: 10.1037/a0028605 [DOI] [PubMed] [Google Scholar]
  • 53.Mills ML. Impairment, Disability, and Substance Use Disorder. Subst Use Misuse. 2023;58(2):221–228. doi: 10.1080/10826084.2022.2155475 [DOI] [PubMed] [Google Scholar]
  • 54.Wollman SC, Hauson AO, Hall MG, et al. Neuropsychological functioning in opioid use disorder: A research synthesis and meta-analysis. Am J Drug Alcohol Abuse. 2019;45(1):11–25. doi: 10.1080/00952990.2018.1517262 [DOI] [PubMed] [Google Scholar]
  • 55.Melugin PR, Nolan SO, Siciliano CA. Bidirectional causality between addiction and cognitive deficits. Int Rev Neurobiol. 2021;157:371–407. doi: 10.1016/bs.irn.2020.11.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Benítez-Silva H, Buchinsky M, Man Chan H, Cheidvasser S, Rust J. How large is the bias in self-reported disability? J Appl Econom. 2004;19(6):649–670. doi: 10.1002/jae.797 [DOI] [Google Scholar]
  • 57.Ledingham E, Adams RS, Heaphy D, Duarte A, Reif S. Perspectives of adults with disabilities and opioid misuse: Qualitative findings illuminating experiences with stigma and substance use treatment. Disabil Health J. 2022;15(2, Supplement):101292. doi: 10.1016/j.dhjo.2022.101292 [DOI] [PubMed] [Google Scholar]
  • 58.Reif S, Karriker-Jaffe KJ, Valentine A, et al. Substance use and misuse patterns and disability status in the 2020 US National Alcohol Survey: A contributing role for chronic pain. Disabil Health J. 2022;15(2, Supplement):101290. doi: 10.1016/j.dhjo.2022.101290 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Brener ND, Billy JOG, Grady WR. Assessment of factors affecting the validity of self-reported health-risk behavior among adolescents: evidence from the scientific literature. J Adolesc Health. 2003;33(6):436–457. doi: 10.1016/S1054-139X(03)00052-1 [DOI] [PubMed] [Google Scholar]
  • 60.Harrison LD, Martin SS, Enev T, Harrington D. Comparing Drug Testing and Self-Report of Drug Use Among Youths and Young Adults in The General Population: (637122007–001). Published online 2007. doi: 10.1037/e637122007-001 [DOI] [Google Scholar]
  • 61.Hunter D 2017 Methodological Summary and Definitions. Substance Abuse and Mental Health Services Administration; 2018. Accessed August 5, 2025. https://www.samhsa.gov/data/report/2017-methodological-summary-and-definitions [Google Scholar]
  • 62.Jackson CT, Covell NH, Frisman LK, Essock SM. Validity of Self-Reported Drug Use Among People with Co-Occurring Mental Health and Substance Use Disorders. J Dual Diagn. 2005;1(1):49–63. doi: 10.1300/J374v01n01_05 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Tables 1-2
NIHMS2208287-supplement-Supplementary_Tables_1-2.docx (32.9KB, docx)

RESOURCES