“It’s all connected:” A mixed methods study of insomnia, stigma, and discrimination among individuals on medication for opioid use disorder

Abstract

Objectives:

Insomnia is one of the most common sleep disorders among those with opioid use disorder (OUD), including those on medication for OUD. There is a dearth of literature exploring the role of social stressors on sleep outcomes among this group. The purpose of this study was to explore the association between OUD-related stigma and intersectional discrimination with insomnia among individuals on medication for OUD.

Methods:

Participants were recruited from treatment clinics in the Northeast United States. Using a convergent mixed-methods research design, we explored associations with stigma (The Brief Opioid Stigma Scale), intersectional discrimination (Intersectional Discrimination Index), and insomnia (Insomnia Severity Index) through quantitative survey data and qualitative data from interviews for participant experiences. Data from the quantitative (n = 120) and qualitative (n = 25) components of the study were integrated for interpretation.

Results:

Quantitative analysis indicated weak to moderate positive correlations between intersectional discrimination, and exploratory variables including pain, perceived stress, and psychological distress with insomnia severity. The qualitative analysis generated 4 main themes, which highlighted negative emotions and ruminations as factors that participants connected experiences with stigma and discrimination to poor sleep outcomes. Integration of data identified concordant and discordant findings.

Conclusions:

Stigma, discrimination, physical symptoms, and psychological distress appear to contribute to poor sleep outcomes among those with OUD. Future research should target maladaptive outcomes of rumination and negative emotions to improve sleep outcomes among those with OUD.

Introduction

Opioid use disorder (OUD) is a chronic and relapsing brain disease defined as the persistent use of opioids despite adverse consequences.¹ OUD affects over 2.7 million individuals in the United States (US), and its prevalence increases annually.¹ OUD often leads to deficits in cognitive function, mood, pain perception, and autonomic activity¹; at least one overdose death every 20 minutes³; and even death.² Despite the availability of medication for opioid use disorder (MOUD), an effective federally approved treatment, only 20% of individuals with OUD receive MOUD each year.⁴ Among those who seek treatment, relapse rates are high and are associated with a lack of retention in treatment. This leads to poor physical, mental, emotional, and behavioral health.³

A major health issue among those with OUD during both active disease and recovery is insomnia,^3,5 defined as difficulty initiating sleep, maintaining sleep, or awakening too early in the morning.⁶ Insomnia occurs in as many as 75% of the people with OUD,³ and individuals who take MOUD report more severe insomnia symptoms than those who do not.⁷ Factors contributing to insomnia in this population include chronic drug use, pain, anxiety, depression, withdrawal symptoms, benzodiazepine use, and cigarette smoking.⁸

OUD is associated with stigma and discrimination. OUD-related stigma is a social process characterized by negative stereotyping, excluding, rejecting, blaming, or devaluing a person based on their OUD. Discrimination is the act of making unjustified, prejudiced distinctions between people as a result of stigma.⁹ Both stigma and discrimination contribute to negative opioid-related outcomes.⁹ People on MOUD typically identify with more than one stigmatizing identity (eg, a person of color, a person with a history of mental illness, a person with a history of criminal justice involvement) that intersect in an inseparable manner that causes “disadvantage by multiple sources of oppression”^10,11 or intersectional discrimination.¹² For example, a person of color with OUD and a history of criminal justice involvement may experience discrimination due to their criminal history, race, and OUD.

Stigma and discrimination are associated with symptoms of insomnia, such as daytime tiredness or sleepiness, nighttime awakenings, shorter sleep duration, and the need to use sleep aids (eg, dietary supplements) to fall asleep or improve sleep quality.^13,14 However, there is a gap in the literature using an intersectional lens to explore stigma, discrimination, and insomnia among those on MOUD. The purpose of this mixed-methods study was to gain a comprehensive understanding of the relationship between stigma and discrimination and insomnia among individuals on MOUD by integrating quantitative and qualitative data. The aims were (1) to describe the association among OUD-related stigma, intersectional discrimination, and exploratory variables (pain, perceived stress, psychological distress) with insomnia among those on MOUD; (2) Describe how individuals on MOUD perceive stigma, discrimination, and sleep.

The conceptual underpinnings of this study were drawn from three overarching frameworks: an adapted version of the health stigma and discrimination framework (see Fig. 1),^13,15 intersectionality,¹² and the allostatic model of discrimination, sleep, and cardiometabolic risk.¹⁶ Blending these frameworks incorporates the intersecting, and multi-factorial drivers and facilitators of stigma and discrimination and the processes by which experiences may manifest and lead to poor sleep quality.

Fig. 1.

The adapted health stigma and discrimination framework.¹³

Methods

To improve understanding of variable associations and participant experiences, we used a convergent parallel mixed methods design (see Fig. 2) that integrates benefits from quantitative and qualitative methods.¹⁷ This study was part of a larger study, collaboration linking opioid use disorder and sleep (CLOUDS) (an ongoing HEAL-funded study [U01HL150596]), which was designed to examine the biological, behavioral, social, and environmental factors that explain the relationship between sleep disturbance, relapse, and retention with MOUD. For our sub-study, we collected quantitative and qualitative data and analyzed them simultaneously.¹⁷ Individuals were enrolled in the CLOUDS study and recruited from low-barrier community treatment centers in Connecticut that provide MOUD. To recruit participants, we used medical record screening by research assistants to identify eligible individuals to contact, word-of-mouth, and recruitment flyers posted in clinics. Individuals who attended the participating clinics completed computer-based surveys lasting approximately 60 minutes at the time of their clinic visits and received $30 for the completion of baseline surveys.

Fig. 2.

A convergent parallel mixed methods design.¹⁵

After individuals completed the questionnaires, we interviewed them to gain additional information about topics in the surveys, including discrimination, stigma, and sleep. We approached 25 adults, purposively sampled with respect to the diversity of insomnia severity index (ISI) scores, gender, age, race, and ethnicity, to participate in interviews. All individuals we approached agreed to participate. The first author (U.N.) conducted face-to-face semi-structured interviews in a private room in the clinics. All interview questions were pilot-tested before interviews (see Table 1). The interpretive descriptive approach, a qualitative methodology that emphasizes the generation of clinically relevant knowledge, was used to address the qualitative research aim.^18,19 Participation was voluntary, and individuals received $40 as compensation. The Yale University IRB approved the study. All individuals provided written informed consent prior to engaging in any study procedures for both the quantitative and qualitative components.

Table 1.

Interview guide.³¹

Semi-structured interview guide for qualitative phase
Introduction:
	Thank you so much for agreeing to participate in this study, we truly appreciate your time. The purpose of this interview is better understand who you are as a person, and your experiences of stigma and discrimination (if any), and how these experiences have impacted you, and your health (including your sleep). For reference, we define stigma as labeling, stereotyping, exclusion, rejection, blame or devaluation of an individual because of attributes viewed as different from societal norms and discrimination as differential or unfair treatment. Also, as a reminder, the questions I will be asking you are personal and can be sensitive in nature, so if you are to ever feel uncomfortable with answering a question, or need to stop, or take a break, please feel free to let me know. Do you have any questions before we get started?
1.	Perceived stigma:
	Prompt:
	a. How do you think people view those with OUD/on treatment for OUD?
	b. Some people feel like other people look down on them—has that been something you’ve experienced—Yes/No—tell me more about that.
2.	Experienced stigma:
	Prompt:
	a. Have you had any direct experiences of stigma or discrimination?—Yes/No—tell me more about that.
	b. Have these experiences been related to OUD and/or being on treatment? Can you say more about that, or give me an example?
	c. What are other characteristics that you feel are an important part of you, but people treated you differently because of? For instance, this can be things such as your race, culture, religion, your job.—Can you say more about that, or give me an example??
	d. What are your experiences with these combined identities? *Repeat participants response to previous question* Do you think it has an impact on how people treat you—yes/no—tell me more about that?
	e. Do you feel like one identity causes you to experience stigma or discrimination more than another identity...if so? Which one? Why?
3.	Internalized stigma:
	Prompt:
	c. The way that other people think of drug users....how does that make you feel about yourself?
	d. How do your experiences of stigma/discrimination make you feel about yourself?–Has it changed the way you think or feel about yourself?
	e. How do your experiences with your combined identities make you feel about yourself?
4.	Anticipated stigma:
	Prompt:
	a. Do you tell people about your OUD and treatment status? How do the people you tell react?
	b. What do you think others’ attitudes are (or would be) toward you if they knew about your OUD or treatment status?—Do you think it would change how they treat you? In what ways?
5.	Sleep:
	Prompt: Thank you for sharing and talking about your experiences with stigma and discrimination. I want to shift gears a bit here and talk about your sleep.
	a. How would you describe your sleep?
	b. What makes you sleep well?
	c. What makes you not sleep well?
	d. Do you feel your experiences with discrimination have contributed to poor night’s sleep? In what ways?
6.	Is there anything I have missed or anything else you would like to tell me? Is there anything that you wished I would have asked that I did not?

OUD, opioid use disorder.

Variables and measures

Demographic and clinical characteristics

Age, race, ethnicity, sex, income, education, health status, employment status, and housing status were collected via self-report. We considered psychological distress, perceived stress, and pain as exploratory variables. All variable measures demonstrated good internal consistency in the present study (α = .83-.91).

Psychological distress

The brief symptom inventory-18 constitutes three domains,–anxiety, depression, and somatic symptoms—with 6 questions for each domain. The global severity index of distress indicates the sum of the 3 domains ranging from 0–72.²⁰

Perceived stress

The perceived stress scale is a self-report questionnaire that consists of 10 items designed to indicate how unpredictable, uncontrollable, and overloaded respondents find their lives.²¹ Scores 0–13 indicate no or mild perceived stress, 14–26 indicate moderate perceived stress and 27–40 indicate severe perceived stress.²²

Pain

Pain was assessed with the brief pain inventory,²³ which has 11 items, 4 for pain severity (at its worst, at its least, on average, and right now), each rated on a 0–10 scale from no pain to pain as bad as you can imagine and pain interference (0–10 scale, does not interfere to completely interferes).²³ Severity scores are the sum of the severity items divided by 4; interference scores are the sum of the interference items divided by 7.²³

Stigma

To measure OUD-related stigma, we use the brief opioid stigma scale,²⁴ which consists of 12 items assessing 4 opioid-related stereotypes across 3 Subscales. It assesses perceived stigma, or stereotype awareness (“aware”); internalized stigma, or stereotype agreement (“agree”); and self-esteem decrement (“harm”) related to opioid dependence. Each subscale score ranges from 4–20.

Intersectional discrimination

The intersectional discrimination index measures day-to-day intersectional discrimination that occurs on the basis of more than 1 perceived characteristic, such as gender, ethnicity, and mental health diagnosis.²⁵ The main scale is formed by 9 Likert-type items with 4 response options (1 never—4 many times).²⁵

Insomnia symptoms

Insomnia symptoms were measured using the ISI,²⁶ a seven-item, self-report questionnaire. Scores range from 0–28, with higher scores indicating higher levels of insomnia; a score of 0–7 indicates no insomnia, 8–15 indicates subthreshold insomnia, 15–21 indicates moderate insomnia, 22–28 indicates severe insomnia.²⁶ An ISI cutoff score of 15 was used for identifying clinical insomnia. We performed exploratory analysis using an ISI cutoff score of 11 to maximize specificity and sensitivity to detect insomnia.⁴⁷

Objective sleep characteristics

In the CLOUDS cohort, objective sleep data were collected using an Actigraph, a wrist-worn device that contains an accelerometer that measures rest-activity patterns and objective sleep variables (eg, sleep onset latency, wake after sleep onset, early morning awakening, total wake time, total sleep time, and sleep efficiency).³² Participants wore actigraphs continuously for 2 weeks. Participants completed sleep diaries, which were used to assist with scoring the actigraph data in Actiware v.6 software (Respironics Minimitter, Inc).

Data management

Quantitative data

Data were analyzed with R software.²⁷ Summary statistics (ie, mean, medians, standard deviations) were computed for all primary study variables. The CLOUDS cohort consisted of 130 individuals at the time of data analysis; those with incomplete data for the primary variables (OUD-related stigma, intersectional discrimination, and insomnia) or exploratory variables (perceived stress, pain, and psychological distress) were excluded from the final analyses (n = 10), for a final sample size of 120. Medians were used where the data were skewed for OUD-related stigma, intersectional discrimination, pain, stress, and psychological distress variables. These variables also did not satisfy assumptions of normality (tested by the Shapiro–Wilk statistic). Spearman correlations were used to assess associations between OUD-related stigma, intersectional discrimination, stress, pain, and psychological distress with insomnia. We compared the means of variables between people with and without insomnia based on their ISI scores using independent group t-tests. We performed all tests at the significance level of 0.05.

Qualitative data

The first author conducted in-person semi-structured interviews and audio-recorded them. Interviews lasted for approximately 45 minutes. Audio recordings were sent to a Health Insurance Portability and Accountability Act-certified transcriptionist. All transcribed interviews were read in their entirety for accuracy. Qualitative data from interview transcripts, field notes, and memos were managed with Atlas.ti Version 8.4.4 (Berlin, Germany). Data coding and analysis occurred simultaneously with interviews until thematic informational redundancy was noted.^28,29 Three co-authors (U.N., M.O., L.S.) developed a coding framework and refined it via consensus. After an 80% inter-coder agreement was reached on jointly coded transcripts, U.N. coded the remaining transcripts. Thematic analysis, a method used to analyze qualitative data that emphasizes classifying and interpreting patterns of meaning, was used to identify key themes and sub-themes.^18,19,30

Quantitative and qualitative data were analyzed separately and then merged.¹⁷ We maintained a spreadsheet containing quantitative data and qualitative themes for each participant. Data integration occurred by comparing survey results with qualitative themes. We illustrated these comparisons in a joint display that presented quantitative and qualitative data concordance, discordance, and interpretations.

Reflexive statement

It is critical for researchers to practice active reflexivity, or the act of acknowledging how one’s experiences, assumptions, and beliefs can influence the research process.¹⁷ The first author, U.N., conducted all interviews; was one of the coders; and kept a diary that enabled her to document bias and record significant events, feelings, and participant interactions. U.N. has clinical experience working with people who use drugs and MOUD. She identifies with commonly stigmatized identities that have led to discriminatory events. These experiences helped develop trust and rapport with participants and facilitated rich dialog. To limit bias in data analysis, interviews were coded by 2 additional coders experienced in qualitative analysis, mixed methods, and community-engaged research.

Results

Quantitative results

The sample consisted of 71 men and 49 women (M age = 41.58 years, SD = 11.64). The majority identified as non-Hispanic White (79%), followed by Hispanic or Latino (17.5%), multi-racial (10.8%) and non-Hispanic Black (8.3%). Most participants had high school diplomas, General Educational Developments or post-high school education (75%) and were unemployed (69.2%). All individuals were being treated for opioid addiction with methadone. Table 2 contains descriptive statistics.

Table 2.

Demographic and clinical characteristics of participants (N = 120).

	Overall sample (N = 120) Mean (SD) or N (%)	No clinically significant insomnia (N = 85) Mean (SD) or N (%)	Clinically significant insomnia (N = 35) Mean (SD) or N (%)	P value
Age	41.6 (11.6)	40.6 (11.2)	44.1 (12.5)	.2048
Female sex	49 (40.8%)	29 (34.1%)	20 (57.1%)	.0250
Race				.5738
Black or African-American	10 (8.3%)	6 (7.1%)	4 (11.4%)
White	95 (79.2%)	69 (81.2%)	26 (74.3%)
More than one race	13 (10.8%)	8 (9.4%)	5 (14.3%)
Unknown	2 (1.7%)	2 (2.4%)	0
Hispanic	21 (17.5%)	14 (16.5%)	7 (20.0%)	.7919
Educationa				.7808
Eighth grade or less	2 (1.7%)	1 (1.2%)	1 (2.9%)
Some high school	22 (18.3%)	14 (16.5%)	8 (22.9%)
High school graduate/GED	48 (40.0%)	34 (40.0%)	14 (40.0%)
Post-high school	42 (35.0%)	30 (35.3%)	12 (34.3%)
Any employment	37 (30.8%)	27 (31.8%)	10 (28.6%)	.8294

GED, General Educational Development.

^ahas missing data.

Using the ISI²⁴ cut-off score of 15, 29% percent of the sample met the criteria for clinical insomnia. Individuals were divided into two groups (no clinically significant insomnia/clinically significant insomnia).²⁶ The descriptive characteristics of the 2 groups are displayed in Table 2. Individuals with clinically significant insomnia were slightly older compared to those without clinically significant insomnia (mean age 44.06 SD = 12.46 vs. 40.55 SD = 11.20). Most individuals with clinically significant insomnia were White (74.3%), and slightly more than half were women (57.1%). The average pain, stress, and psychological distress scores were higher among those with clinically significant insomnia compared to those with no clinically significant insomnia (Table 3). Similar findings were noted in using the cut-off score of 11 (Supplementary Table 1).

Table 3.

Descriptive statistics: pain, stress, intersectional discrimination, OUD-related stigma, psychological distress.

	Overall sample (N = 120)		No clinically significant insomnia (N = 85)		Clinically significant insomnia (N = 35)		P value	Cohen’s d
Variable/measure	Median	Mean (SD)	Median	Mean (SD)	Median	Mean (SD)
Pain severity (BPI)	2.5	2.89 (2.70)	1.75	2.36 (2.58)	4.5	4.19 (2.60)	.0008	0.71
Pain interference (BPI)	1.93	2.95 (3.05)	1	2.30 (2.85)	4.86	4.51 (3.00)	.0010	0.75
Perceived stress (PSS)	21	21.11 (4.54)	20	20.47 (4.48)	22	22.66 (4.36)	.0121	0.49
Intersectional discrimination (IDI)	3	4.91 (5.33)	2	4.27 (4.75)	4	6.46 (6.34)	.0753	0.39
Stigma awareness (BOSS_Aware)	17.5	16.51 (3.93)	17	16.07 (4.15)	19	17.57 (3.16)	.0654	0.41
Stigma agreement (BOSS_Agree)	9	10.03 (4.38)	10	10.13 (4.63)	8	9.80 (3.76)	.8732	0.06
Self-esteem decrement (BOSS_Harm)	7	7.54 (4.20)	6	7.32 (4.16)	7	8.09 (4.29)	.3125	0.18
Somatization (BSI)	50	53.43 (9.23)	50	51.15 (7.90)	58.5	59.13 (9.95)	.0002	0.89
Depression (BSI)	50	53.76 (10.48)	48	51.19 (9.51)	61	60.19 (10.17)	< .0001	0.91
Anxiety (BSI)	49	52.57 (10.30)	48	50.14 (8.94)	58	58.66 (11.06)	.0002	0.75
Global severity (BSI)	53.5	54.18 (10.30)	49.5	51.29 (9.31)	61	61.41 (9.14)	< .0001	1.09

Bold values denote statistical significance at the p < 0.05 level.

There was a weak positive correlation (r_s = 0.17, P = .04) between intersectional discrimination and insomnia severity. There were no statistically significant correlations between stigma and insomnia severity. Pain severity, pain interference, perceived stress, and psychological distress were moderately correlated with insomnia severity (Table 4).

Table 4.

Bivariate Spearman correlations with study variables and insomnia severity.

Variable/measure	Correlation	P value
Stigma awareness (BOSS_Aware)	0.18	.09
Stigma agreement (BOSS_Agree)	0.05	.58
Self-esteem decrement (BOSS_Harm)	0.09	.33
Intersectional discrimination (IDI)	0.18	.04
Pain severity (BPI)	0.42	< .0001
Pain interference (BPI)	0.45	< .0001
Somatization (BSI)	0.45	< .0001
Depression (BSI)	0.49	< .0001
Anxiety (BSI)	0.41	< .0001
Global_Severity (BSI)	0.46	< .0001
Perceived stress (PSS)	0.21	.01

p, significance level.

Bold values denote statistical significance at the p < 0.05 level.

Qualitative results

The sample consisted of 25 individuals. More than half (14) were women. The mean age was 43.8 (SD = 12.41) years. Most identified as non-Hispanic White (76%), 24% identified as non-Hispanic Black, and 20% identified as Hispanic or Latino. Table 5 displays the demographic characteristics of the qualitative sample.

Table 5.

Demographic and clinical characteristics of participants (N = 25).

Baseline characteristics	Enrolled participant (N = 25) Mean (SD) or N (%)
Demographics
Age at enrollment	43.8 (12.41)
Sex
Male	11 (44.0%)
Female	14 (56.0%)
Ethnicity
Hispanic or Latino	5 (20.0%)
Not Hispanic or Latino	20 (80.0%)
Race
Black or African American	6 (24.0%)
White	19 (76.0%)
Where does the participant live?
Own home/apartment	19 (76.0%)
Assisted living	2 (8.0%)
Homeless	2 (8.0%)
Other	2 (8.0%)
With whom is the subject living
Alone	5 (20.0%)
With spouse or partner	10 (40.0%)
With family other than spouse	5 (20.0%)
With other	5 (20.0%)
Education level
8th grade or less	1 (4.0%)
9th to 11th grade	4 (16.0%)
High school diploma or GED	9 (36.0%)
Vocational trade school courses or associates degree courses after high school	2 (8.0%)
Courses toward a bachelors degree	7 (28.0%)
Bachelors degree	1 (4.0%)
Masters degree	1 (4.0%)
Employment status
Full-time	5 (20.0%)
Part-time	2 (8.0%)
Unemployed, looking for work	12 (48.0%)
Unemployed, not looking for work	4 (16.0%)
Retired	1 (4.0%)
Unable to work	1 (4.0%)

GED, General Educational Development; SD, standard deviation.

The analysis resulted in 4 main themes, each with respective sub-themes: (1) living with multiple Identities and addiction, (2) it’s everywhere: discrimination and stigma, (3) perceptions about sleep, and (4) connecting discrimination and stigma with sleep. Themes 1 and 2 are reported briefly below and have been fully described elsewhere (manuscript under review).³¹

Living with multiple identities and addiction

All but one individual described identities that intersected with OUD addiction and treatment stigma.³¹ Identities were associated with characteristics, affiliations, and social roles, specifically race, gender, age, comorbid conditions (ie, mental illness, chronic pain, physical disability), physical appearance (ie, teeth, hair color, scars from drug use), single parenthood, low socioeconomic status, work roles (ie, sex work, adult dancing), and having had criminal justice involvement. The most common intersecting identities were (1) a person of color with OUD, (2) low socioeconomic status background with OUD, and (3) criminal justice involvement with OUD. When reflecting on multiple identities, one individual stated,

...It’s unfortunate and just me, my disabilities and my skin tone and my drug use, I think those are the major things that I really feel like people have treated me badly because of includin’ the people that’s supposed to love you. It’s been tough.

It’s everywhere: discrimination and stigma

All individuals recalled negative biases and stereotypes against people with OUD: “They think the people that do drugs ain’t s * ** . They don’t want nothin’ in life. They’re bums, as far as they’re concerned....” Sources of stigma and discrimination included family, friends, community members, co-workers and employers, government services (ie, child services, bus, banks, housing), drug treatment centers/detox clinics, the criminal justice system, and medical institutions.³¹

Participants often internalized discriminatory and stigmatizing experiences, resulting in feeling shame, guilt, embarrassment, worthlessness, damage, and/or incapable. Repeated experiences of being stereotyped led some individuals to begin to believe the stereotypes as true: “I think, personally—through all the experiences that I’ve been through [with stigma and discrimination], I self-sabotage. Because I believe everything that’s being said about me, that I’m not good enough for anything.” ³¹

In contrast, some individuals denied negatively internalizing experiences because they were unconcerned about what people thought, or they were not ashamed of who they were. For example, one individual stated, “Yeah, I know what they all think, but I’m confident in who I am. I don’t let what people say get to me...it doesn’t define me.”³¹.

Perceptions about sleep

Individuals varied on how they perceived their sleep quality. Many described sleep as “erratic,” “very bad” and something one “could not get enough of.” Few individuals described their sleep positively and as something they looked forward to doing: “I love my sleep, my sleep is my vacation.”

Disrupted sleep patterns.

Most individuals indicated difficulty either falling asleep (ie, taking hours to fall asleep) or staying asleep. Some discussed nocturnal awakenings due to feeling restless, needing to use the restroom, or because that was what their body was accustomed to: “I wake up every two hours on the dot...my body just wakes up every two hours.” They identified daytime sleepiness as a consequence of disrupted sleep patterns, which often led to frequent napping as a compensatory strategy. While many used napping as a way to “catch up” on their sleep time, a few described not sleeping for days at a time, commonly related to stress and worrying: “I’ve been just so stressed out with life and everything going on. I don’t sleep at all. At this point I have to go days without sleeping to exhaust myself so I can at least sleep a little bit...it’s terrible.”

Participants described several factors that affected sleep.

including racing minds, sleep disorders (insomnia, sleep apnea), inactivity, sleeping too much, unstable living conditions, worrying, noise (neighborhood noise), drug use (caffeine, opioids), stress, waking up during the night, negative emotions (heartbreak, feeling like a disappointment), electronic use (lights, tv), discomfort in bed, having a “long day,” sleeping on a full stomach, and pain. The most cited factors included racing minds, stress, drug use, or a combination of two or more.

I sleep okay. It’s been a little bit hard recently because I just been going through so much with my family, just stress with COVID. I’ve lost some loved ones. Some financial issues. It has been tough, so I haven’t been sleeping that well. I think I only sleep about three hours a night because of everything that I got going on. I got to get work done. My mind is just wanderin,’ so stressed out about everything that’s goin’ on.

Factors that supported sleep were over-the-counter-sleep medications, illicit drugs (ex: marijuana, opioids), reading, listening to white noise, exercising, keeping busy during the day/exhausting oneself during the day, having a warm drink before bed, following a sleep/wake schedule, relaxing, having a stuffed animal in the bed, and having basic life necessities (food, shelter, clothing, money) and for a few, sleeping during daylight due to neighborhood safety concerns. Stable living conditions facilitated better sleep as described by one individual: “Just having a bed, a roof over my head, food, and yeah. Pretty steady cashflow of money coming in right now. That helps. That’s what helps me sleep well at night.it makes me feel at ease.”

Connecting discrimination and stigma to sleep

Individuals often expressed emotions such as shame, guilt, and feeling unworthy resulting from stigmatizing and discriminatory experiences. They perceived these emotions as contributing to sleep problems, such as delayed sleep onset, waking during the night, or not sleeping at all. In some cases, people simply did not allow their sleep to be disrupted or used sleep to avoid the pain of stigma and discrimination. One individual who described being discriminated against often due to her multiple stigmatized identities (mentally challenged, criminal justice involvement, recovering addict) stated:

Sometimes I get into those thoughts where I start to believe what has been said to me through the years or the looks, and you know what those looks are. What they mean, and you start questioning yourself, “Am I an able-bodied person? Am I a good person? Am I the one to blame for everything happening to me?” I question myself sometimes, and sometimes it’s not—I’m not happy with the answer. It makes me sad. It kind of gives me the heebie-jeebies and causes me to keep myself up at night with my thoughts.

Individuals described rumination, (ie, continuously thinking negative thoughts) at night. Rumination connected their stigmatizing and discriminatory experiences to their sleep. Many individuals described replaying discriminatory experiences repeatedly in their head during the night. When one individual reflected on their rumination, they stated,

...I feel like I can be quite sensitive regarding all of this [discrimination/stigma], it stresses me out. The judgment and the stress really gets me. It keeps me up at night because I keep thinking over and over and over again how unfairly people mistreat and judge me for who I am. My mind just doesn’t rest.

“Sleep is my escape.”

Some individuals made few or no overt connections between stigma, discrimination and sleep. For some, the lack of connection was a result of having few or no experiences with stigma and discrimination. Others used sleep to block out problems and negative thoughts related to stigma and discrimination. One individual stated, “I don’t let any of that stuff [discrimination/stigma] get to me. My sleep is important to me, so I don’t let anything affect it. Once I hit the pillow, I knock out completely.” Another individual even went on to describe how they felt sleep was their way to “escape” their reality and the hurt and pain that came with being stigmatized and discriminated against.

Integration

To gain a more comprehensive understanding of the association between stigma, discrimination, and insomnia in individuals on MOUD, we integrated the quantitative and qualitative data analysis and findings¹⁷ in 2 phases. The first phase included comparing the larger quantitative sample findings with the qualitative sub-sample findings. The second phase included merging the results of the quantitative and qualitative data to compare and contrast and identify data concordance and discordance. A color-coded joint display of integrated data is represented in Supplementary Fig. 1.

Sleep characteristics

The scores on the ISI indicated that most individuals had no clinically significant insomnia (37.5% had no insomnia, 33.3% subthreshold, 20% moderate, 9.2% severe). In contrast, qualitative findings demonstrated that most individuals perceived their sleep quality as very poor, and many experienced several symptoms of insomnia (eg, waking up during the night, difficulty falling asleep at night).

In the parent study, sleep data were collected with actigraphy.³² Figs. 3 and 4 used selected actigraph data from the parent study to illustrate examples of discordance between self-report and objectively measured sleep. Fig. 3 illustrates actigraph data from a participant who reported greater insomnia severity, but whose actigraphy report demonstrated ‘healthy’ sleep through regular sleep and wake times, no major movements while sleeping, and regular activity during awake hours. Fig. 4 depicts actigraph data from a participant who did not subjectively report insomnia but had actigraphy data consistent with insomnia symptoms, such as irregular sleep and wake patterns.

Fig. 3.

Actigraphy report of participant exemplar case—self-report insomnia w/ healthy actigraphy sleep. Note. This is an actigraphy report for a 39-year-old Puerto Rican female, with self-reported insomnia. Notably, she described her sleep quality as poor in the qualitative component of the study. Features of this actigraphy data report include an 8-day period starting from 12 PM and ending at 11:59 AM, the activity level indicated by the black and red line under each day, the yellow line which indicates light (natural or artificial), light blue shaded intervals which indicate intervals of rest, and dark blue shades intervals which indicate intervals of sleep.

Fig. 4.

Actigraphy report of participant exemplar case—self-report no insomnia w/ poor actigraphy sleep. Note. This is an actigraphy report for a 35-year-old Black male, with no self-reported insomnia, that described their sleep quality as poor in the qualitative component of the study. Features of this actigraphy data report include an 8-day period starting from 12 PM and ending at 11:59 AM, the activity level indicated by the black and red line under each day, the yellow line which indicates light (natural or artificial), light blue shaded intervals which indicate intervals of rest, and dark blue shades intervals which indicate intervals of sleep.

The average score for OUD-related stigma awareness was high, as was the prevalence in qualitative findings of participants discussing their awareness of public stigma related to drug addiction and treatment. Although quantitative scores indicated that stigma agreement and self-esteem decrement were relatively low, qualitative data suggested that repeated exposure to stereotypes eventually led to stigma agreement and may have negatively affected overall self-esteem by reinforcing negative feelings and emotions.

Scores on the intersectional discrimination index were relatively low. This was discordant with the qualitative data, as many participants reported living with 2 or more stigmatizing identities resulting in discrimination. Discriminatory experiences were more harmful to the overall well-being for individuals who held multiple minoritizing identities since they were more likely to experience recurring and chronic discrimination.

Connecting discrimination, stigma, and sleep

The 3 subscales of OUD-related stigma were not significantly correlated with insomnia severity. However, qualitative data highlighted that the OUD-related stigma fostered negative feelings, often leading to restlessness and rumination that affected their ability to fall and stay asleep. The weak but positive correlation between intersectional discrimination and insomnia was confirmed by the qualitative findings: more than half of the participants reported a connection between their discrimination experiences and troubled sleep.

Connecting pain, stress, psychological factors, and sleep

Quantitatively, significant moderate correlations were found between perceived stress, psychological distress, and pain with insomnia severity. The qualitative data supported these correlations: many psychological and physiological symptoms were described as barriers to participants’ sleep quality and quantity.

Discussion

To our knowledge, this is the first in-depth investigation of the association among OUD-related stigma, intersectional discrimination, and insomnia among individuals on MOUD. We found a weak positive correlation between intersectional discrimination and insomnia severity, while there were moderate positive correlations for pain, stress, and psychological distress with insomnia severity. This was further supported by qualitative analysis of interview data in which participants described perceived stigma, discrimination, and sleep and ways in which they felt that stigma and discrimination affected their sleep. The mixed-methods approach broadened our understanding of stigmatizing, discriminatory, and sleep experiences among individuals on MOUD and strengthened the validity of the findings.

Most individuals did not have clinically significant insomnia, based on responses to the self-report questionnaire. Despite these reports, interview data revealed poor sleep quality and many symptoms associated with insomnia. This discrepancy may be explained in several ways, as outlined by Harvey and Tang (2012).³³ Individuals with insomnia may misperceive their sleep by overestimating the time it takes to get to sleep and underestimating the total sleep time, and/or they may experience psychological distress causing magnification of symptoms.³³ For example, Hughes et al. (2018) found that underserved older adults had discordant findings between self-reported sleep (measured using the Pittsburgh sleep quality index and ISI) and actigraphy-measured sleep.³⁴ Discrepancies in sleep survey scores and interview data may also be due to the notable chronicity of sleep problems in this population; sleep problems may be dismissed or normalized and regarded as consequences of drug use and treatment, leading to under-reporting on questionnaires versus interviews. Individuals may also not perceive poor sleep to be as bothersome when compared to other daily and life stressors, such as financial instability. Discrepancies may also be attributed to the use of our cutoff score of 15 on the ISI. Optimal cutoff scores on the ISI can vary from one population to another. Therefore, we performed exploratory analysis using an ISI cutoff score of 11 (optimal in detecting insomnia in community samples⁴⁷) to ensure specificity and sensitivity to identify insomnia in this population.⁴⁷ However, the results did not differ based on this categorization (Supplementary Table 1).

Scores on the 3 stigma subscales provided insight into the level of stigma awareness, stigma agreement, and self-esteem among participants. Although scores were lower than anticipated overall, scores for stigma awareness were the highest of the 3 subscales, consistent with qualitative data. Individuals were aware of stigma and often felt misunderstood and prejudged. Similar results were found in a qualitative study that sought to explicate experiences of stigma among methadone maintenance therapy patients.³⁵ Half of the participants reported experiencing at least some negative stereotypes associated with methadone maintenance therapy, and this contributed to decreased self-esteem and guilt.³⁵ Our participants repeatedly reported internalizing negative experiences with stigma and discrimination and perceived many stereotypes and prejudices aimed at them.

Although many studies have documented various forms of stigma relating to substance use and treatment status,^36–38 we identified various identities that intersected with substance use stigma beyond those that are most commonly studied: race/ethnicity, gender, sexual orientation, and socioeconomic status. Most stigmatizing experiences described by participants in our study resulted from intersecting experiences with OUD, race/ethnicity, socioeconomic background, and/or criminal histories. In interviews, many individuals identified multiple identities that intersected with their OUD and recalled hurtful discriminatory and stigmatizing experiences based on these intersecting identities. Inconsistencies between survey data and qualitative findings often exist in studies of minoritized or underserved populations,^39,40 possibly due to sensitivity of measures of discrimination or participants intentionally under-reporting discrimination experiences as a way of coping with feelings of rejection or managing pain associated with discriminatory experiences.^39,40

Our finding that OUD-related stigma was not significantly correlated with insomnia severity contrasts with research that found positive correlations between stigma and sleep.¹³ Individuals on MOUD may have perceived more discrimination or been stigmatized for other identities, such as being a person with a history of criminal justice involvement or poorer socioeconomic background more than their addiction and treatment status. Individuals often maintain a positive appearance and conceal their engagement with treatment programs to avoid disclosing their OUD to others,⁴¹ which could result in experiencing OUD stigma and discrimination.⁴¹ Many individuals described the importance of keeping a “clean and neat” appearance to avoid negative labeling and stereotyping.

Finally, intersectional discrimination, pain, perceived stress, and psychological distress were positively and significantly correlated with insomnia severity in our quantitative findings; this was supported by our qualitative data. Overall, our findings align with prior research suggesting that stigma and discrimination affect various health outcomes, including sleep and are often associated with physical symptoms (eg, pain), and psychological symptoms (eg, anxiety, depression, stress). ^13,14

Participants identified rumination and negative emotions as mechanisms by which discriminatory and stigmatizing experiences may affect sleep. Evidence suggests that in rumination, repetitive thoughts fuel physiological, cognitive, and emotional arousal, which disrupts sleep, leading to insomnia.^42,43 A cross-sectional study among a sample of 68 African Americans found a significant indirect effect of racial discrimination on participants’ sleep quality through rumination.⁴⁴ Another study investigated the association between psychosocial factors and self-reported sleep duration and quality in a racially and ethnically diverse sample of 1326 adults and found that rumination moderated the relationship between race/ethnicity discrimination and subjective sleep quality and duration. However, contrary to our findings, anxiety and depression symptoms were not consistently associated with these sleep outcomes.⁴⁵

Strengths and limitations

The strengths of this study include its mixed methods design, which yields a robust and comprehensive description of this complex clinical problem. Also, we worked to maximize rigor and trustworthiness by using valid and reliable instruments for quantitative data collection and, for the qualitative analysis, by engaging in reflexivity and using multiple team members for coding and analytic processes.^17,46

Several limitations should be considered. The cross-sectional nature of this study precludes examination of causal relationships among study variables, and the quantitative sample lacked diversity in type of treatment medication, geographical location, race, and ethnicity. Although the quantitative sample had adequate statistical power, the sample size was small, limiting our analysis approach to correlational analysis, which produces less conclusive results than more robust techniques. The small sample may have also contributed to the decreased statistical power and effect size of the correlations. Although survey measures used to assess OUD-related stigma, intersectional discrimination, and insomnia were psychometrically sound, reporting periods were limited to specific time frames (eg, past 2 weeks), and survey instructions and language were complex, insensitive, and lacked description; this may have left participants needing further clarification. However, the qualitative interview in our study successfully elicited a range of experiences of sleep and intersectional discrimination by fostering a safe and comfortable environment for individuals to discuss sensitive matters, defining and giving examples of poor sleep outcomes and intersectional discrimination to clarify the concept, and encouraging conversation through sensitive and responsive probing and prompting.

Study implications

Future prospective studies with larger and more diverse samples are needed for more sophisticated analyses to examine the impact of stigma and discrimination on sleep outcomes among those with addiction. More information is needed about associations between stigma and discrimination, and objective sleep characteristics and their impact on stages of sleep (awake, light, deep, and rapid eye movement sleep), which each play a unique role in mental and physical health. Future work developing scales for discrimination, stigma, and sleep should include input from both experts and the target population, to ensure that instruments are sensitive, understandable, and psychometrically sound. Future research should aim to identify intervention approaches that target rumination and other psychological processes to help mitigate the effects of discrimination and stigma on self-perception to improve sleep outcomes for people with OUD and other intersecting identities.

Our findings have important implications for clinicians and other health providers working with individuals with OUD. It is important for providers to understand how stigma and discrimination disrupt sleep so that they can suggest sleep-enhancing alternatives, such as behavioral and psychosocial interventions to promote sleep health and overall well-being.

Conclusion

In this convergent mixed methods study, we explored the relationships between OUD-related stigma, intersectional discrimination, and insomnia among patients on MOUD; how individuals perceive stigma, discrimination, and sleep; and how individuals believe their experiences with discrimination and stigma are linked to sleep. Although some data from the quantitative and qualitative components of the study were discordant, our integrated quantitative and qualitative findings contribute to the body of evidence linking intersectional discrimination with physical symptoms, psychological symptom distress, and insomnia within this highly stigmatized population. Participants reported that rumination and negative emotions contribute to insomnia, which suggests that interventions to improve sleep should consider the contributions of stigma, discrimination, rumination, and negative emotions to improve sleep outcomes among those with OUD.

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.sleh.2023.09.004.