Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Apr 1.
Published in final edited form as: J Pain. 2023 Oct 30;25(4):984–999. doi: 10.1016/j.jpain.2023.10.019

Baseline Characteristics from a New Longitudinal Cohort of Patients with Non-cancer Pain and Chronic Opioid use in the United States

Scott Secrest 1, Lisa R Miller-Matero 4, Timothy Chrusciel 1,3,6, Joanne Salas 1,3, Mark D Sullivan 5, Celeste Zabel 4, Patrick Lustman 7, Brian Ahmedani 4, Ryan W Carpenter 8, Jeffrey F Scherrer 1,2,3
PMCID: PMC10960712  NIHMSID: NIHMS1941170  PMID: 37907114

Abstract

Retrospective cohort studies have consistently observed that long-term prescription opioid use is a risk factor for new major depressive episodes. However, prospective studies are needed to confirm these findings and to establish evidence for causation. The Prescription Opioids and Depression Pathways cohort study is designed for this purpose. The present report describes the baseline sample and associations between participant characteristics and odds of daily vs. non-daily opioid use. Second, we report associations between participant characteristics and odds of depression, dysthymia, anhedonia, and vital exhaustion. Patients with non-cancer pain were eligible if they started a new period of prescription opioid use lasting 30 to 90 days. Participants were 54.8 (SD±11.3) years of age, 57.3% female and 73% white race. Less than college education was more common among daily vs. non-daily opioid users (32.4% vs. 27.3%; p=0.0008), as was back pain (64.2% vs. 51.3%; p<0.0001), any non-opioid substance use disorder (12.8% vs. 4.8%; p<0.0001) and current smoking (30.7% vs. 18.4% p<0.0001). High pain interference (50.9% vs. 28.4%; p<0.0001) was significantly associated with depression, as was having more pain sites (6.9±3.6 vs. 5.7±3.6; p<0.0001), and benzodiazepine co-mediation (38.2% vs. 23.4%; p<0.0001). High pain interference was significantly more common among those with anhedonia (46.8% vs. 27.4%; p<0.0001) and more pain sites (7.0±3.7 vs. 5.6±3.6; p<0.0001) were associated with anhedonia. Having more pain sites (7.9±3.6 vs. 5.5±3.50; p<0.0001) was associated with vital exhaustion as was back pain (71.9% vs. 56.8%; p=0.0001) and benzodiazepine co-medication (42.8% vs. 22.8%; p<0.0001). Patients using prescription opioids for non-cancer pain have complex pain, psychiatric, and substance use disorder comorbidities. Longitudinal data will reveal whether long-term opioid therapy leads to depression or other mood disturbances such as anhedonia and vital exhaustion.

Keywords: pain, opioid, psychiatry, cohort, epidemiology, mood

INTRODUCTION

A growing body of evidence indicates that long-term opioid therapy (LTOT), independent of pain and comorbid psychiatric disorders, is associated with approximately twice the risk for new depression episodes as compared to short term opioid use.1,2 A Mendelian Randomization analysis found evidence for a causal relationship between LTOT and risk for depression and anxiety.3 However, these studies only measure diagnosed depression in general and it is possible that LTOT is associated with depression-like conditions including dysthymia (i.e. chronic mild depression), anhedonia or vital exhaustion. Anhedonia is the inability to experience pleasure and vital exhaustion is characterized by loss of energy, demoralization and irritability.4–6

To our knowledge, there are three existing prospective cohort studies that have measured the association between prescription opioid use and new or worsening depression. In a primary care cohort with non-cancer pain, dose escalation was associated with worsening depression.7 In two other prospective cohort studies, opioid therapy was not associated with depression. Von Korff and colleagues8 suggest that their older patient population and intermittent, mostly low-dose opioid users could account for a lack of association between LTOT and development of depression. The Pain and Opioids IN Treatment (POINT) study9 investigators observed an association between prescription opioid use and new onset depression, but this was not significant after controlling for pain duration and comorbid psychiatric disorders. Of note, the median duration of opioid use was 4 to 6 years in the POINT study,9,10 which is longer than the 3.4 years previously observed between starting a new period of opioid use and onset of depression.2 This could have limited their ability to detect an association LTOT and new onset depression. Additionally, recent evidence indicates only daily or near daily opioid use is associated with risk for depression among those with LTOT.11 Therefore, more studies of factors associated with daily vs. intermittent use may help explain why opioids are associated with mood disturbance. The mechanism for the relationship between LTOT and new onset depression is unknown. One possibility is that long-term occupation of opioid receptors leads to anhedonia. Neurophysiological changes in areas of the brain associated with reward have been observed in patients with LTOT,12,13 and this may contribute to onset of anhedonia and major depression. Sleep disturbance14,15, hyperalgesia16, and androgen deficiency,17 are potential consequences of LTOT that could contribute to increased risk for depression. The prospective Prescription Opioids and Depression Pathways cohort study, henceforth termed “Pathways”18 was created in part to determine if LTOT leads to incident major depressive or actually leads to other forms of disturbed mood, such as dysthymia, anhedonia or vital exhaustion. and second to identify opioid exposures most strongly associated with mood disturbance. Last, the study will identify moderators and mediators of the LTOT -mood disturbance association.

The present descriptive study uses baseline data to explore associations between pain measures, opioid exposures and participant comorbidities most strongly associated with daily vs. non-daily opioid use. Second, we explored the associations between the same exposure variables and major depression, dysthymia, anhedonia and vital exhaustion. By identifying which pain and opioid use measures are most strongly associated with these phenotypes, we will inform future research by narrowing the mood disturbances linked to non-cancer pain and near daily or daily prescription opioid use. Last, results provide the first description of the Pathways cohort at baseline.

METHODS

Details about the Pathways study protocol have been previously published.18 In brief, patients were eligible for Pathways if they were a patient at Saint Louis University’s academic medical practice, St. Louis, Missouri or a patient at Henry Ford Health, Detroit, Michigan. Eligible adult (18-80 years of age) participants were identified from the electronic health record (EHR) on a weekly basis. Eligible patients were free of cancer and cancer treatment per EHR data, and potential participants confirmed cancer status via screening questions prior to interview. All patients had started a new period of prescription opioid use lasting 30 to 90 days at baseline. We defined a new period of opioid use as having no opioid prescription in the past 3 months. The eligibility criteria for 30-90 day baseline opioid use were applied to enrich the sample with participants likely to transition to long-term prescription opioid therapy (LTOT) defined as >90 days of opioid use We used screening questions to confirm that patients had been prescribed a new opioid medication within the last 3-months and were not being treated for cancer. Study packets were mailed to potential participants and contained instructions on how to participate and informed consent materials. A respondent booklet was sent to participants to enable survey completion. To keep participant responses contemporary with their new period of prescription opioid use, we allowed 5 weeks to complete the survey after mailing study packets. Therefore, some participants had >90 days of opioid use by the time they completed the baseline survey. To participate, patients had to release their EHR data from the 12 months prior to baseline and for the 12 months of follow-up. After providing informed consent, participants either completed the survey on their own (66.4%) by logging into the REDCap survey instrument or by calling a project research assistant (33.6%) who then entered answers into REDCap.

The baseline survey was pre-tested and revised until it took an average of 60 minutes to complete. Participants were given a $50 gift card for each survey and were eligible for a $1,000 raffle if all survey items were completed at 6-month follow-up and again at 12-month follow-up. Baseline enrollment began 11/2019 and 1,047 participants were enrolled at the end of baseline, 11/2022. Sampling and recruitment outcomes are illustrated in e-Figure 1.

All study procedures were reviewed and approved by the Saint Louis University and Henry Ford Health IRBs.

Survey measures:

Demographic questions included age, race, gender, ethnicity, employment status, educational attainment, household income and marital status.

We used the Brief Pain Inventory19,20 to measure pain severity and pain interference. Pain severity was measured via 4-items: worst in last 30 days, least in last 30 days, pain on average, and current pain. Pain severity was the average of these 4-items on a scale from ‘0=no pain’ to ‘10=pain as bad as you can imagine’. High pain severity was a score of 7 or above. Seven pain interference questions assessed whether pain interfered with general activity, mood, walking ability, normal work, relationships, sleep, and enjoyment of life in the last 30 days. The pain interference score was the average of these 7-items on a scale of ‘0=does not interfere’ to ‘10=completely interferes’. High pain interference was defined as a score of ≥8. The number of pain locations was obtained by asking participants if they had pain in 17 distinct body locations (e.g., lower back, arms, foot etc.). Specific pain conditions were obtained from EHR data. Using ICD-10 codes, we measured the following pain diagnoses: arthritis, back/neck pain, musculoskeletal pain, neuropathies, migraine/headache, fibromyalgia, and chronic pain.

Opioid prescription characteristics were measured with the following questions: 1) type of opioid (e.g., codeine, dihydrocodeine, fentanyl, hydrocodone, etc. prescribed ) prescribed in last 90 days; 2) daily opioid dose (using unit dose and number taken per day), which was converted to morphine milligram equivalent dose (MME) and MME dose was dichotomized as < 50 MME per day vs. ≥50 MME per day and 3) frequency of use was defined as daily or not daily. The Prescribed Opioids Difficulty Scale (PODS) measured psychosocial problems that patients attributed to opioid use and patient concerns about opioid use. Higher scores indicate greater problems with opioids, including side effects and perceived misuse symptoms. Scores ≥16 were considered a high overall PODS score.21

Non-opioid pain therapies were obtained by survey by asking if participants ever tried cannabis, chiropractic care or physical therapy. EHR data was used to measure receipt of chiropractic treatment, interventional pain management, physical therapy, and prescriptions for NSAIDs, gabapentin and muscle relaxants. Each type of treatment was counted and modeled as the number of non-opioid treatments measured by survey and separately measured from EHR data.

We used a computerized version of the Semi-Structured Assessment for the Genetics of Alcoholism (SSAGA-II)22 to obtain DSM-IV lifetime and past year depression, dysthymia, bipolar disorder, opioid use disorder and any non-opioid substance use disorder. The Generalized Anxiety Disorder scale-7 (GAD-7) screened for generalized anxiety disorder (GAD), with a score of ≥10 indicating at least moderate anxiety.23 The Primary Care-PTSD-5 (PC-PTSD-5) was used to screen for posttraumatic stress disorder (PTSD),24 with a score of ≥3 indicating probable PTSD. The Snaith-Hamilton Pleasure Scale (SHAPS)25 was used to measure anhedonia with higher scores indicating more severe anhedonia and a score ≥ 3 indicating high anhedonia.26. The Maastricht Vital Exhaustion brief form assessed vital exhaustion which is a measure of “unusual fatigue, increased irritability and feelings of demoralization”.5 Higher scores indicate worse vital exhaustion with a score of ≥10 indicating high vital exhaustion.5 Hazardous alcohol use was measured with the AUDIT-C screener, with a score of ≥3 for women and ≥4 for men indicating hazardous drinking.27 Lifetime smokers were those who reported ever smoking 100 or more cigarettes in their lifetime. Based on current smoking status (“do you now smoke every day, some days, or not at all?”), participants were classified as never/past vs. current smokers.

EHR data was used to measure specific types of pain diagnoses, mental health treatments received including prescriptions for any antidepressant and any benzodiazepine as well as any psychotherapy visits. The Charlson comorbidity index was created using EHR data on chronic health problems.28–30 Higher scores indicate worse morbidity and risk for mortality.

The PROMIS SF v2.0–Emotional Support 4a scale assessed emotional support with higher scores indicating more social support.31 The PROMIS Ability to Participate in Social Roles and Activities v.2 measured perceived barriers to leisure, family, work and activity with friends. Barriers are not anchored to pain.32 The PROMIS social roles and emotional support scales were scored as “high” if the T-scaled score based on a published reference population was ≥60. Last, the Pittsburgh Sleep Quality Index measured sleep disorders including difficulty falling asleep, waking, discomfort due to temperature and pain and fatigue and sleepiness.33 Global scores ranged from 0 to 21 with higher scores indicating poorer sleep quality.

Weighting data for non-participation

All analyses were weighted using stabilized inverse probability of participation weights so that results could be generalized to all eligible patients who had non-cancer pain and a new period of 30-90 days of opioid use. Propensity scores (PS) for participation in the Pathways Study were calculated using a binary logistic regression model assessing the conditional participation based on age, race, gender, and electronic health record variables (arthritis, back/neck pain, muscle pain, fibromyalgia, chronic pain, neuropathy, headache, any substance use disorder, depression, and anxiety). Stabilized weights were calculated by multiplying the inverse of the propensity score by the observation participation rate. Stabilized weights reduce bias associated with extreme rates and preserve original sample size in analyses.34–36 Assessment of balance for included model variables was calculated using the standardized mean difference percent (SMD%). As shown in appendix, e-table 1 and e-table 2, all variables balanced between those participating and not participating in the study as all SMD% were <10%.37 E-table 3 shows the distribution of study variables before and after weighting. Successful weighting indicates that results are generalizable to eligible patients with non-cancer pain, starting a new period of 30-90 day opioid use.

Analytic Approach

Prior to conducting analyses, all data was weighted to represent the total eligible population. The weighted sample size after applying inverse probability of participation weights was 1,039. First, univariate measures of distribution are presented for selected cohort characteristics. For continuous characteristics, this includes the cohort mean and standard deviation (SD). All continuous characteristics were evaluated and met the criteria for normality based on skewedness and kurtosis. For categorical characteristics percent and count in each category are reported. The bivariate associations between cohort characteristics and each of the following outcome variables were computed: daily opioid use, past year depression, past year dysthymia, anhedonia, and vital exhaustion. Bivariate association between cohort characteristics and each key variable was assessed using Chi-square and Fisher’s Exact tests for categorical variables and t-tests for continuous variables. SMD% was used as a measure of effect size for bivariate comparisons, and as mentioned previously, an SMD%>10% was used to indicate meaningful differences. All analyses were weighted using the methodology described above. Because of the large number of comparisons, we used a conservative alpha of 0.001 all statistical tests. SAS v9.4 (Cary, NC) was used for all analyses.

RESULTS

As shown in Table 1, participants were an average of 54.8 years of age (SD = 11.3), 73.0% were white, 57.3% were women, 51.0% were currently unemployed or on disability due to pain or another reason, 52.2% were married or living with a partner, and 29.7% had an annual household income of less than $20,000.

Table 1.

Pathways Study participants characteristics overall at baseline (n=1047) weighted3

Demographics % (n) or mean (SD)
Age (mean, SD) 54.8 (11.3)
Gender
  Man 42.5% (442)
  Woman 57.3% (596)
  Missing <5
Race
  White 73.0% (759)
  Black 21.9% (228)
  Other 2.2% (23)
  Missing 2.8% (29)
Ethnicity
  Hispanic or Latino 2.8% (29)
  Not Hispanic or Latino 95.2% (989)
  Missing 2.1% (22)
Employment
  Working now 25.5% (266)
  No work now 19.8% (206)
  Disability due to pain 19.8% (206)
  Disability other reason 11.4% (118)
  Retired 22.7% (236)
  Missing 0.7% (8)
Education
  No College 30.4% (316)
  Some College 46.9% (488)
  Bachelor or higher 22.5% (234)
  Missing <5
Household income
  <$20,000 29.7% (309)
  $20,000-$49,999 24.2% (251)
  $50,000-99999 23.6% (245)
  100K + 16.3% (169)
  Missing 6.3% (66)
Marital Status
  Married/live with partner 52.2% (543)
  Widow/Div/Sep 29.8% (309)
  Never married 16.9% (175)
  Missing 1.1% (12)
Pain measures
High pain severity
  Yes 28.7% (298)
  No 71.2% (740)
  Missing <5
High pain interference
  Yes 33.0% (343)
  No 66.9% (696)
  Missing <5
BPI # pain sites—mean (SD) 5.9 (3.7)
EHR pain measures
Arthritis
  Yes 65.9% (685)
  No 34.1% (355)
Back pain
  Yes 59.6% (620)
  No 40.4% (420)
Musculoskeletal pain
  Yes 54.8% (570)
  No 45.2% (470)
Neuropathies
  Yes 15.9% (165)
  No 84.1% (875)
Migraine/headache
  Yes 13.9% (145)
  No 86.1% (895)
Fibromyalgia
  Yes 6.5% (67)
  No 93.5% (973)
Chronic Pain not elsewhere classified
  Yes 44.0% (457)
  No 56.0% (582)
Prescription opioid measures
Daily opioid use
  Yes 66.1% (688)
  No 31.5% (327)
  Missing 2.4% (25)
Daily MME ≥50 mg
  Yes 15.3% (159)
  No 73.2% (761)
  Missing 11.5% (120)
High PODs score
  Yes 16.8% (174)
  No 82.0% (852)
  Missing 1.3% (13)
Non-opioid therapies (from survey)1
  0 7.9% (82)
  1 31.5% (328)
  2 38.5% (400)
  3 22.0% (229)
  Missing <5
Non-opioid therapies from EHR2
  0 14.7% (153)
  1 15.9% (166)
  2 21.8% (226)
  3 22.9% (238)
  4 16.6% (173)
  5 7.8% (82)
  6 <5
Psychiatric conditions
Depression past year
  Yes 20.3% (211)
  No 79.5% (827)
  Missing <5
Depression lifetime
  Yes 31.8% (331)
  No 68.1% (708)
  Missing <5
Dysthymia past year
  Yes 6.8% (71)
  No 93.0% (967)
  Missing <5
Dysthymia lifetime
  Yes 15.7% (163)
  No 84.1% (875)
  Missing <5
Bipolar past year
  Yes 2.3% (23)
  No 97.6% (1015)
  Missing <5
Bipolar lifetime
  Yes 4.9% (51)
  No 95.0% (987)
  Missing <5
Anhedonia
  Yes 28.3% (294)
  No 69.9% (727)
  Missing 1.8% (19)
Vital Exhaustion
  Yes 18.0% (187)
  No 81.3% (845)
  Missing 0.8% (8)
GAD positive
  Yes 19.4% (201)
  No 80.1% (833)
  Missing 0.6% (6)
PTSD positive
  Yes 15.3% (159)
  No 82.1% (853)
  Missing 2.6% (27)
AUDIT-C positive
  Yes 20.1% (209)
  No 72.5% (754)
  Missing 7.4% (77)
Opioid use disorder past year
  Yes 0.4% (5)
  No 99.1% (1030)
  Missing 0.4% (5)
Opioid use disorder lifetime
  Yes 3.4% (35)
  No 96.2% (1000)
  Missing 0.4% (5)
Any non-opioid SUD past year
  Yes 1.2% (12)
  No 98.6% (1025)
  Missing <5
Any non-opioid SUD lifetime
  Yes 10.1% (105)
  No 89.6% (932)
  Missing <5
Current Smoker
  Yes 26.4% (275)
  No 72.9% (758)
  Missing 0.7% (7)
  Co-medications (from EHR, in past year)
Benzodiazepine
  Yes 26.5% (275)
  No 73.5% (765)
Any antidepressant
  Yes 47.4% (493)
  No 52.6% (547)
Any psychotherapy
  Yes 2.9% (30)
  No 97.1% (1009)
Psychosocial
High emotional support
  Yes 53.1% (552)
  No 46% (478)
  Missing 1% (10)
High Social functioning
  Yes 11.7% (122)
  No 86.5% (899)
  Missing 1.8% (18)
Sleep disorder (PSQI)—mean (SD) 10.1 (4.3)
Comorbidity Index—mean (SD) 1.8 (2.3)
Open in a new tab
1)

survey based measures of CBD use or cannabis use for pain, chiropractic, PT (ever/never for these);

2)

EHR based measures of chiropractic, interventional pain, NSAID, gabapentin, muscle relaxant, PT (past year);

3)

Data weighted to the source population (eligible patients invited to participate). Sum of weights is 1,039

The most common opioid used was hydrocodone (28.5%) followed by tramadol (23.2%), oxycodone (15.8%) and codeine (6.1%). Other opioid medications were uncommon (1.4%) and 25.1% were using more than one type of opioid medication.

The mean number of pain sites was 5.9 (SD = 3.7). One third of participants had high pain interference, and over a quarter had high pain severity. The most prevalent pain condition was arthritis (65.9%) and least prevalent was fibromyalgia (6.5%). At baseline, 66.1% were daily prescription opioid users and 15.3% were taking ≥50 MME a day. Nearly 17% had a high PODs score and most participants had tried 2 or more non-opioid therapies.

The most prevalent psychiatric condition was lifetime major depressive disorder (31.8%) and the least common disorder was lifetime bipolar disorder (4.9%). Approximately 20% of participants had current depression, 19.4% screened positive for GAD and 15.3% screened positive for PTSD. Among substance use variables, 0.4% of respondents were positive for past year opioid use disorder and 3.4% had a lifetime opioid use disorder. 20.1% of respondents screened positive for alcohol use disorder and more than 26.4% were current smokers. More than 1 out of 4 had a benzodiazepine prescription and 47.4% had an antidepressant. Over half had high emotional support but only 11.7% reported high social functioning. The mean Pittsburgh Sleep Quality Index (PSQI) score was 10.1 (SD = 4.3, range = 0-21).

As shown in Table 2, current employment (p<0.0001, SMD%=27.1) and a college degree (p=0.0008, SMD%=24.8) were less common among daily prescription opioid users. High pain interference (p<0.0001, SMD%=27.6) was more prevalent among daily users. Back pain (p<0.0001, SMD%=26.3) was more prevalent among daily users. Higher MME dose (SMD%=51.3), a high PODS score (SMD%=36.8), three self-reported non-opioid pain therapies from the survey (SMD%=21.2) and up to five non-opioid therapies from the EHR (SMD% range = 17.5-34.6) were all significantly associated with daily opioid use compared to non-daily use (p<0.0001). Any lifetime, non-opioid substance use disorder (p<0.0001, SMD%=28.5) and current smoking (p<0.0001, SMD%=28.7) were each more common among daily vs. non-daily users.

Table 2.

Pathways Study participants characteristics by prescription opioid use frequency at baseline

daily opioid use (n=688) non-daily opioid use (n=327) p-value SMD%
Demographics
Age (mean, SD) 54.4 (11.2) 55.8 (11.3) 0.0497 13.2
Gender 0.0588
  Man 44.5% (305) 38.2% (125) 12.8
  Woman 55.5% (381) 61.8% (202) 12.8
Race 0.0205
  White 77.4% (517) 69.9% (222) 17.1
  Black 20.1% (135) 28.1% (89) 18.6
  Other 2.5% (17) 2.1% (7) 2.9
Ethnicity 0.6470
  Hispanic or Latino 3.1% (20) 2.5% (8) 3.2
  Not Hispanic or Latino 97% (652) 97.5% (313) 3.2
Employment <.0001
  Working now 22.1% (150) 34.1% (111) 27.1
  No work now 22.7% (155) 14.2% (46) 22.1
  Disability due to pain 22.2% (152) 15.4% (50) 17.7
  Disability other reason 11.4% (78) 11.4% (37) 0.1
  Retired 21.6% (147) 24.9% (81) 7.9
Education 0.0008
  No College 32.4% (223) 27.3% (89) 11.3
  Some College 48.2% (332) 42.8% (139) 11.0
  Bachelor or higher 19.3% (133) 29.9% (97) 24.8
Household income 0.1932
  <$20,000 33.5% (217) 28.2% (86) 11.6
  $20,000-$49,999 26.4% (171) 24.5% (75) 4.3
  $50,000-99999 23.6% (153) 27.9% (85) 9.8
  100K + 16.4% (106) 19.4% (59) 7.7
Marital Status 0.6855
  Married/live with partner 52.6% (357) 54.2% (176) 3.1
  Widow/Div/Sep 30.9% (210) 28.3% (92) 5.8
  Never married 16.4% (112) 17.5% (57) 2.9
Pain measures
High pain severity 31.4% (216) 24.4% (80) 0.0221 15.6
High pain interference 37.2% (256) 24.6% (80) <.0001 27.6
BPI # pain sites—mean (SD) 6.2 (3.7) 5.5 (3.5) 0.0025 20.6
EHR pain measures
Arthritis 64.8% (446) 70.3% (230) 0.0826 11.8
Back pain 64.2% (441) 51.3% (168) <.0001 26.3
Musculoskeletal pain 57.5% (395) 49.9% (163) 0.0231 15.3
 Neuropathies 16.5% (113) 14.5% (47) 0.4198 5.5
 Migraine/headache 14.5% (100) 13.5% (44) 0.6757 2.8
Fibromyalgia 7.0% (48) 5.9% (19) 0.5078 4.5
Chronic Pain 45.8% (315) 40.7% (133) 0.1257 10.3
Prescription opioid measures
Daily MME ≥50 mg 22.8% (141) 5.5% (16) <.0001 51.3
High PODS score 21.6% (146) 8.6% (28) <.0001 36.8
Non-opioid therapies (from survey)1 .003
  0 6.2% (42) 10.8% (35) 16.7
  1 30.9% (213) 33.0% (108) 4.4
  2 37.7% (259) 39.6% (129) 3.9
  3 25.2% (173) 16.6% (54) 21.2
Non-opioid therapies from EHR2 <.0001
  0 10.2% (70) 23.2% (76) 35.4
  1 12.4% (85) 22.9% (75) 27.9
  2 19.3% (133) 26.7% (87) 17.5
  3 27.1% (187) 14.2% (47) 32.3
  4 20.9% (144) 8.8% (29) 34.6
  5 9.8% (67) 4.2% (14) 22.2
  6 <5 <5 7.1
Psychiatric conditions
Depression past year 21.8% (150) 17.7% (58) 0.1252 10.5
Depression lifetime 34.1% (234) 27.0% (88) 0.0234 15.5
Dysthymia past year 7.9% (54) 4.7% (15) 0.0585 13.3
Dysthymia lifetime 17.5% (120) 12.5% (41) 0.0419 14.0
Bipolar past year 2.7% (18) 1.5% (5) 0.2547 8.0
Bipolar lifetime 5.7% (39) 3.6% (12) 0.1524 10.0
Anhedonia 29.0% (196) 29.1% (93) 0.9676 0.3
Vital Exhaustion 19.1% (130) 16.9% (55) 0.4099 5.6
GAD positive 20.5% (140) 17.6% (57) 0.2685 7.5
PTSD positive 16.5% (111) 14.2% (45) 0.3741 6.1
AUDIT-C positive 21.6% (138) 22.2% (67) 0.8525 1.3
Opioid use disorder past year <5 <5 0.6702 4.7
Opioid use disorder lifetime 4.3% (30) 1.6% (5) 0.0250 16.3
Any non-opioid SUD past year 1.4% (9) <5 0.4097 3.8
Any non-opioid SUD lifetime 12.8% (88) 4.8% (16) <.0001 28.5
Current Smoker 30.7% (209) 18.4% (60) <.0001 28.7
  Co-medications
Benzodiazepine 28.2% (194) 23.5% (77) 0.1158 10.7
Any antidepressant 47.9% (330) 46.7% (153) 0.7185 2.4
Any psychotherapy 3.2% (22) 2.5% (8) 0.5257 4.4
Psychosocial
High emotional support 52.6% (359) 54.6% (176) 0.5393 4.2
High Social functioning 10.1% (68) 15.0% (48) 0.0225 15.0
Sleep disorder (PSQI)—mean (SD) 10.3 (4.4) 9.8 (4.3) 0.1756 9.6
Comorbidity Index—mean (SD) 1.7 (2.2) 2.1 (2.4) 0.0370 13.9
Open in a new tab
1)

survey based measures of CBD use or cannabis use for pain, chiropractic, PT

2)

EHR based measures of chiropractic, interventional pain, NSAID, gabapentin, muscle relaxant, PT

As shown in Table 3, participants with past year depression at baseline were younger than those without depression (p<0.0001, SMD%=40.5). High pain interference (p<0.0001,SMD%=47.3), and a greater number of pain sites (p<0.0001, SMD%=34.3) were all significantly associated with past year depression, as was a high PODs score (p<0.0001, SMD%=42.5). All psychiatric covariates, except for hazardous drinking, opioid use disorder in the last year or lifetime, and current smoking, were related to past year depression at p<.0001 (SMD% range=26.2-101.8). Current prescription for a benzodiazepine (p<0.0001, SMD%=32.4) or antidepressant (p<0.0001, SMD%=36.6) were more common among those with past year depression. Less prevalent high emotional support, high social functioning, and worse sleep quality were more common in those with previous year depression at (p<0.0001, SMD% range=35.9-77.9).

Table 3.

Pathways Study participants characteristics by past year depression at baseline

Past Year Depression (n=211) No Past Year Depression (n=827) p-value SMD%
Demographics
Age (mean, SD) 51.2 (11.3) 55.8 (11.1) <.0001 40.5
Gender 0.0027
  Man 33.5% (71) 45.0% (371) 23.6
  Woman 66.5% (140) 55% (455) 23.6
Race 0.3365
  White 76.3% (154) 74.9% (605) 3.4
  Black 20.2% (41) 23.1% (187) 7.1
  Other 3.5% (7) 2.0% (16) 9.0
Ethnicity 0.0648
  Hispanic or Latino 4.7% (10) 2.3% (19) 12.9
  Not Hispanic or Latino 95.3% (198) 97.7% (790) 12.9
Employment 0.2352
  Working now 22.2% (47) 26.6% (218) 10.4
  No work now 23.0% (48) 19.2% (157) 9.4
  Disability due to pain 23.0% (48) 19.1% (157) 9.6
  Disability other reason 12.5% (26) 11.2% (92) 4.0
  Retired 19.3% (41) 23.8% (196) 11.1
Education 0.3704
  No College 28.2% (60) 31.0% (256) 6.2
  Some College 45.7% (97) 47.3% (390) 3.2
  Bachelor or higher 26.1% (55) 21.7% (179) 10.4
Household income 0.0201
  <$20,000 40.2% (79) 29.5% (229) 22.7
  $20,000-$49,999 24.8% (49) 26.1% (202) 3.0
  $50,000-99999 22.5% (44) 25.8% (201) 7.9
  100K + 12.5% (25) 18.6% (144) 16.7
Marital Status 0.0419
  Married/live with partner 45.4% (94) 54.8% (449) 18.9
  Widow/Div/Sep 33.5% (69) 29.3% (240) 9.2
  Never married 21.1% (44) 15.9% (130) 13.3
Pain measures
High pain severity 34.3% (73) 27.3% (225) 0.0429 15.3
High pain interference 50.9% (107) 28.4% (234) <.0001 47.3
BPI # pain sites—mean (SD) 6.9 (3.6) 5.7 (3.6) <.0001 34.3
EHR pain measures
Arthritis 69.8% (147) 64.8% (536) 0.1791 10.5
Back pain 65.2% (138) 58.2% (481) 0.0661 14.3
Musculoskeletal pain 57.1% (121) 54.1% (448) 0.4441 5.9
 Neuropathies 20.9% (44) 14.6% (120) 0.0243 16.7
 Migraine/headache 19.2% (40) 12.5% (103) 0.0121 18.4
Fibromyalgia 10.5% (22) 5.4% (45) 0.0069 19.0
Chronic Pain 49.2% (104) 42.6% (352) 0.0870 13.2
Prescription opioid measures
Daily OAU 72.2% (150) 66.6% (537) 0.1252 12.1
Daily MME ≥50 mg 15% (28) 17.9% (132) 0.3431 8.0
High PODS score 30.7% (64) 13.5% (110) <.0001 42.5
Non-opioid therapies (from survey)1 .001
  0 7.3% (15) 8.0% (66) 2.5
  1 21.9% (46) 34.0% (281) 27.2
  2 39.9% (84) 38.2% (316) 3.4
  3 30.8% (65) 19.8% (163) 25.7
Non-opioid therapies from EHR2 0.2909
  0 15.1% (32) 14.7% (121) 1.2
  1 12.2% (26) 16.8% (139) 13.2
  2 24.4% (51) 21.1% (175) 7.7
  3 20.7% (44) 23.5% (195) 6.7
  4 20.6% (44) 15.6% (129) 13.0
  5 6.6% (14) 8.1% (67) 5.6
  6 <5 <5 5.2
Psychiatric conditions
Dysthymia past year 15.8% (33) 4.5% (37) <.0001 38.3
Dysthymia lifetime 33.6% (71) 11.1% (92) <.0001 56.1
Bipolar past year 7.7% (16) 0.9% (7) <.0001 34.1
Bipolar lifetime 12.7% (27) 3% (24) <.0001 36.8
Anhedonia 55% (114) 22.1% (180) <.0001 71.8
Vital Exhaustion 51.4% (107) 9.7% (80) <.0001 101.8
GAD positive 47.2% (100) 12.3% (101) <.0001 82.6
PTSD positive 37.9% (78) 10% (80) <.0001 69.2
AUDIT-C positive 25.6% (50) 20.8% (160) 0.1406 11.6
Opioid use disorder past year <5 <5 0.1439 5.5
Opioid use disorder lifetime 5.5% (12) 2.8% (23) 0.0497 13.7
Any non-opioid SUD past year 4.3% (9) <5 <.0001 26.2
Any non-opioid SUD lifetime 17.2% (36) 8.4% (69) 0.0002 26.7
Current Smoker 31.9% (67) 25.3% (208) 0.0512 14.8
  Co-medications
Benzodiazepine 38.2% (81) 23.4% (194) <.0001 32.4
Any antidepressant 61.6% (130) 43.7% (361) <.0001 36.6
Any psychotherapy 6.4% (13) 2% (17) 0.0009 21.7
Psychosocial
High emotional support 38.7% (81) 57.4% (471) <.0001 38.1
High Social functioning 3.9% (8) 14% (114) <.0001 35.9
Sleep disorder (PSQI)—mean (SD) 12.6 (3.7) 9.5 (4.3) <.0001 77.9
Comorbidity Index—mean (SD) 1.6 (2.0) 1.9 (2.4) 0.0537 14.2
Open in a new tab
1)

survey based measures of CBD use or cannabis use for pain, chiropractic, PT

2)

EHR based measures of chiropractic, interventional pain, NSAID, gabapentin, muscle relaxant, PT

Depression, bipolar disorder, anhedonia, vital exhaustion and screening positive for GAD were significantly more common in those with vs. without dysthymia (all p<.0001, SMD% range=38.0-73.6). No other participant characteristics were significantly associated with dysthymia.

As shown in Table 5, high interference (p<0.0001, SMD%=41.1), and a greater number of pain sites (p<0.0001, SMD%=39.0) were all associated with anhedonia. Depression, bipolar disorder, dysthymia, vital exhaustion, and screening positive for GAD or PTSD were significantly more prevalent among those with vs. without anhedonia (p<0.0001, SMD% range=24.3-69.9). Anhedonia at baseline was correlated with lower prevalence of high emotional support (p<0.0001, SMD%=59.2), lower prevalence of high social functioning (p=0.0002, SMD%=28.8) and lower sleep quality (p<0.0001, SMD%=67.9).

Table 5.

Pathways Study participants characteristics by anhedonia at baseline

Anhedonia (n=294) No Anhedonia (n=727) p-value SMD%
Demographics
Age (mean, SD) 54.7 (11.0) 54.9 (11.4) 0.7848 1.9
Gender 0.9729
  Man 42.5% (125) 42.4% (308) 0.2
  Woman 57.5% (169) 57.6% (418) 0.2
Race 0.9707
  White 75.8% (220) 75.3% (529) 1.2
  Black 22.1% (64) 22.4% (157) 0.7
  Other 2.1% (6) 2.3% (16) 1.5
Ethnicity 0.7559
  Hispanic or Latino 2.5% (7) 2.8% (20) 2.2
  Not Hispanic or Latino 97.5% (281) 97.2% (691) 2.2
Employment 0.0005
  Working now 17.4% (51) 28.9% (208) 27.4
  No work now 24.9% (73) 18% (130) 16.8
  Disability due to pain 24.1% (70) 18.3% (132) 14.1
  Disability other reason 12.6% (37) 11.1% (80) 4.7
  Retired 21% (61) 23.7% (171) 6.5
Education 0.5861
  No College 30% (88) 30.3% (220) 0.7
  Some College 45.4% (134) 47.9% (347) 5.1
  Bachelor or higher 24.7% (73) 21.8% (158) 6.8
Household income 0.0018
  <$20,000 35.4% (99) 30.3% (206) 11.0
  $20,000-$49,999 31% (87) 23.8% (162) 16.3
  $50,000-99999 22% (61) 26% (177) 9.4
  100K + 11.6% (32) 20% (136) 23.2
Marital Status 0.0859
  Married/live with partner 47.5% (139) 55.1% (395) 15.2
  Widow/Div/Sep 32.9% (96) 29% (208) 8.6
  Never married 19.5% (57) 16% (115) 9.4
Pain measures
High pain severity 34% (100) 26.4% (192) 0.0149 16.6
High pain interference 46.8% (137) 27.4% (199) <.0001 41.1
BPI # pain sites—mean (SD) 7.0 (3.7) 5.6 (3.6) <.0001 39.0
EHR pain measures
Arthritis 70.5% (207) 63.4% (461) 0.0318 15.0
Back pain 64.5% (190) 58.5% (425) 0.0754 12.4
Musculoskeletal pain 58% (171) 53.4% (388) 0.1865 9.2
 Neuropathies 16.6% (49) 15.7% (114) 0.7317 2.4
 Migraine/headache 14.0% (41) 14.0% (102) 0.9920 0.1
Fibromyalgia 7.3% (22) 6% (44) 0.4390 5.2
Chronic Pain 44.5% (131) 44.2% (321) 0.9269 0.6
Daily OAU 67.8% (196) 67.9% (480) 0.9676 0.3
Daily MME ≥50 mg 16.5% (43) 18% (115) 0.6056 3.8
High PODS score 20.5% (60) 15.6% (112) 0.0586 12.9
Non-opioid therapies (from survey)1 .502
  0 8.3% (24) 7.9% (58) 1.3
  1 29.41% (86) 32.4% (235) 7.1
  2 37.6% (111) 38.7% (281) 2.3
  3 25.0% (74) 21.0% (152) 9.6
Non-opioid therapies from EHR2 0.0103
  0 13.8% (41) 15% (109) 3.5
  1 20.1% (59) 14.2% (103) 15.6
  2 20.9% (61) 22.2% (161) 3.2
  3 16.3% (48) 25.6% (186) 22.8
  4 20.6% (60) 15.1% (109) 14.4
  5 8.1% (24) 7.9% (57) 0.9
  6 <5 <5 3.1
Psychiatric conditions
Depression past year 38.7% (114) 12.8% (93) <.0001 62.0
Depression lifetime 53.9% (159) 22.7% (165) <.0001 67.6
Dysthymia past year 13.3% (39) 4.3% (31) <.0001 32.2
Dysthymia lifetime 28.4% (84) 10.6% (77) <.0001 46.1
Bipolar past year 5.3% (16) 1.1% (8) <.0001 24.3
Bipolar lifetime 9.6% (28) 3.2% (23) <.0001 26.4
Vital Exhaustion 42% (122) 8.6% (62) <.0001 83.4
GAD positive 39.9% (117) 11.1% (81) <.0001 69.9
PTSD positive 26.9% (78) 11.3% (80) <.0001 40.6
AUDIT-C positive 20.9% (56) 22.5% (152) 0.5947 3.8
Opioid use disorder past year <5 <5 0.3777 3.5
Opioid use disorder lifetime 5.7% (17) 2.4% (17) 0.0077 16.9
Any non-opioid SUD past year 1.9% (6) 0.8% (6) 0.3818 9.4
Any non-opioid SUD lifetime 11% (32) 10% (72) 0.6324 3.3
Current Smoker 28.9% (85) 25.5% (185) 0.2706 7.6
  Co-medications
Benzodiazepine 30.2% (89) 25% (182) 0.0926 11.5
Any antidepressant 52.2% (154) 45.7% (332) 0.0625 12.9
Any psychotherapy 5.2% (15) 2.1% (15) 0.0073 16.9
Psychosocial
High emotional support 33.6% (98) 62% (448) <.0001 59.2
High Social functioning 5.8% (17) 14.4% (104) 0.0002 28.8
Sleep disorder (PSQI)—mean (SD) 12.2 (4.3) 9.3 (4.1) <.0001 67.9
Comorbidity Index—mean (SD) 1.9 (2.2) 1.8 (2.3) 0.7605 2.1
Open in a new tab
1)

survey based measures of CBD use or cannabis use for pain, chiropractic, PT

2)

EHR based measures of chiropractic, interventional pain, NSAID, gabapentin, muscle relaxant, PT

As reported in Table 6, high pain severity, high pain interference, and a greater number of pain sites were all significantly associated with vital exhaustion at (p<0.0001, SMD% range=31.1-66.1). Back pain diagnosis (p=0.0001, SMD%=32.0) and fibromyalgia (p=0.0003, SMD%=25.5) were more common among participants with vs. without vital exhaustion. A high PODS score (p<0.0001, SMD%=44.5) and having tried three self-reported non-opioid therapies (p<0.0001, SMD%=32.5) were more common among those with vs. without vital exhaustion. Depression, dysthymia, bipolar disorder, anhedonia, GAD and PTSD positive, and lifetime opioid use disorder were correlated with vital exhaustion at (p<0.0001, SMD% range=28.1-121.8). Compared to those without vital exhaustion, those with vital exhaustion had about double the prevalence of current benzodiazepine prescription and about 1.5 times more the prevalence of any antidepressant (p<.0001, SMD% range=43.7-45.3). More participants with vital exhaustion received psychotherapy compared to those without vital exhaustion (p=0.0001, SMD%=25.4). Vital exhaustion at baseline was correlated with lower prevalence of high emotional support, and high social functioning, and lower sleep quality (all p<0.0001, SMD% range=57.6-111.8).

Table 6.

Pathways Study participants characteristics by vital exhaustion at baseline

Vital Exhaustion (n=187) No Vital Exhaustion (n=845) p-value SMD%
Demographics
Age (mean, SD) 52.7 (10.8) 55.3 (11.3) 0.0042 23.5
Gender 0.0063
  Man 33.8% (63) 44.7% (378) 22.6
  Woman 66.2% (123) 55.3% (467) 22.6
Race 0.1265
  White 78.8% (142) 74.2% (610) 10.8
  Black 17.7% (32) 23.7% (195) 14.9
  Other 3.5% (6) 2.1% (17) 8.9
Ethnicity 0.3908
  Hispanic or Latino 3.6% (7) 2.5% (21) 6.6
  Not Hispanic or Latino 96.4% (179) 97.5% (803) 6.6
Employment 0.0113
  Working now 15.8% (29) 27.9% (233) 29.4
  No work now 23.5% (44) 19.3% (162) 10.3
  Disability due to pain 24.7% (46) 19% (159) 14.0
  Disability other reason 13.2% (25) 11.1% (93) 6.3
  Retired 22.8% (42) 22.8% (191) 0.1
Education 0.6561
  No College 31.7% (59) 30.1% (254) 3.5
  Some College 48.1% (90) 46.7% (393) 2.9
  Bachelor or higher 20.2% (38) 23.3% (196) 7.5
Household income 0.0045
  <$20,000 36.9% (65) 30.3% (240) 14.1
  $20,000-$49,999 31.5% (55) 24.6% (194) 15.5
  $50,000-99999 21.8% (38) 26.0% (205) 9.7
  100K + 9.7% (17) 19.2% (152) 27.2
Marital Status 0.0449
  Married/live with partner 44.8% (82) 54.6% (457) 19.7
  Widow/Div/Sep 36.7% (67) 28.7% (240) 17.1
  Never married 18.5% (34) 16.7% (140) 4.7
Pain measures
High pain severity 40.7% (76) 26.2% (221) <.0001 31.1
High pain interference 57% (107) 27.6% (233) <.0001 62.3
BPI # pain sites—mean (SD) 7.9 (3.6) 5.5 (3.5) <.0001 66.1
EHR pain measures
Arthritis 69.4% (130) 64.9% (549) 0.2441 9.5
Back pain 71.9% (134) 56.8% (480) 0.0001 32.0
Musculoskeletal pain 55.7% (104) 54.6% (461) 0.7930 2.1
 Neuropathies 22.4% (42) 14.4% (122) 0.0064 20.9
 Migraine/headache 20.4% (38) 12.1% (102) 0.0029 22.6
Fibromyalgia 12.2% (23) 5.1% (43) 0.0003 25.5
Chronic Pain 49.7% (93) 42.8% (361) 0.0829 14.0
Daily OAU 70.3% (130) 67.2% (553) 0.4099 6.8
Daily MME ≥50 mg 20.8% (34) 16.5% (124) 0.1903 10.9
High PODS score 31.9% (59) 13.7% (114) <.0001 44.5
Non-opioid therapies (from survey)1 <.0001
  0 4.2% (8) 8.8% (746) 18.6
  1 21.3% (40) 34.0% (287) 28.6
  2 41.0% (77) 37.9% (320) 6.3
  3 33.5% (63) 19.3% (163) 32.5
Non-opioid therapies from EHR2 0.0038
  0 7% (13) 16.4% (138) 29.4
  1 16.7% (31) 15.6% (132) 3.0
  2 24.2% (45) 21.1% (178) 7.4
  3 19.2% (36) 23.9% (202) 11.3
  4 23.9% (45) 15.2% (128) 22.2
  5 8.4% (16) 7.7% (65) 2.6
  6 <5 <5 6.0
Psychiatric conditions
Depression past year 57.3% (107) 11.9% (101) <.0001 108.5
Depression lifetime 72.6% (136) 22.5% (190) <.0001 115.9
Dysthymia past year 15.3% (29) 4.8% (40) <.0001 35.6
Dysthymia lifetime 35.6% (67) 11.2% (95) <.0001 60.2
Bipolar past year 9.1% (17) 0.8% (7) <.0001 39.1
Bipolar lifetime 14.3% (27) 2.9% (25) <.0001 41.4
Anhedonia 66.3% (122) 20.3% (169) <.0001 104.9
GAD positive 60.1% (111) 10.4% (87) <.0001 121.8
PTSD positive 44.5% (81) 9.2% (76) <.0001 86.8
AUDIT-C positive 27.1% (48) 20.7% (162) 0.0607 15.2
Opioid use disorder past year <5 <5 0.3527 1.8
Opioid use disorder lifetime 8.6% (16) 2.3% (19) <.0001 28.1
Any non-opioid SUD past year 3.4% (6) 0.7% (6) 0.0192 18.9
Any non-opioid SUD lifetime 15.5% (29) 9.1% (77) 0.0088 19.7
Current Smoker 33.4% (62) 25.2% (212) 0.0219 18.2
  Co-medications
Benzodiazepine 42.8% (80) 22.8% (192) <.0001 43.7
Any antidepressant 65.2% (122) 43.2% (365) <.0001 45.3
Any psychotherapy 7.3% (14) 2% (17) 0.0001 25.4
Psychosocial
High emotional support 28.3% (52) 59.2% (497) <.0001 65.5
High Social functioning <5 14.6% (122) <.0001 57.6
Sleep disorder (PSQI)—mean (SD) 13.6 (3.5) 9.3 (4.1) <.0001 111.8
Comorbidity Index—mean (SD) 1.9 (1.6) 1.8 (1.6) 0.4556 6.0
Open in a new tab
1)

survey based measures of CBD use or cannabis use for pain, chiropractic, PT

2)

EHR based measures of chiropractic, interventional pain, NSAID, gabapentin, muscle relaxant, PT

DISCUSSION

Baseline data from a new prospective cohort study of patients with non-cancer pain and starting a new period of prescription opioid use revealed that two-thirds were daily opioid users and only 15% had an MME ≥50 mg. Daily as opposed to non-daily opioid users were younger, disproportionately white, unemployed or on disability, and without a college degree. High pain severity and pain interference and more pain locations were positively associated with daily vs. non-daily opioid use.

Pain and prescription opioid use patterns among daily vs. non-daily users were similar to those previously reported from analyses of historical medical record data.38 The high prevalence of unemployment or disability, markedly higher prevalence of MME doses ≥ 50, greater prevalence of lifetime depression and dysthymia and history of opioid use disorder, other forms of SUD and smoking have been previously associated with LTOT.39,40 Daily prescription opioid use is most common in LTOT relative to short-term use, therefore, some of the patient characteristics associated with daily use could be due to these factors occurring frequently among those with LTOT.

To our knowledge, the most similar research design to our own is the POINT study which recruited Australian adults with 6 or more weeks of opioid use for CNCP.10 For instance, we observed 65.9% prevalence of arthritis and 6.5% prevalence of fibromyalgia which is nearly the same as the POINT baseline data (61.6% arthritis and 5.9% fibromyalgia).10 Similarly, the prevalence of PTSD was 15.3% in both the POINT study and our present report, and 22.8% of those in the POINT study had GAD while we found 19.4% with GAD in our cohort. The POINT study observed a 46.6% prevalence of moderate/severe depression and we found 20.3% had a past year major depressive episode. This difference may be due to the POINT study use of the PHQ-9 and our use of a diagnostic interview to measure depression. Overall, the pain and psychiatric comorbidity profiles of patients with LTOT appear mostly consistent between the U.S. and Australia despite the U.S. being by far the largest consumer of prescription opioids in the world.41 The high prevalence of psychiatric comorbidity among patients in the Pathways cohort is consistent with that observed in the POINT study and highlights Campbell and colleague’s conclusion that patients with LTOT have “very complex demographic and clinical profiles.”10

Interestingly, opioid use frequency (daily vs. non-daily) and MME were not significantly associated with past-year depression, dysthymia, anhedonia, or vital exhaustion. This conflicts with evidence that persons with psychiatric disorders receive 50% of all opioid prescriptions dispensed in the U.S.42 However our data are only baseline measures and over follow-up we anticipate daily use and higher doses will be positively associated with new onset depression during the Pathway’s study 12-month follow-up. Patients with depression and vital exhaustion were much more likely than those without these conditions to have a lifetime history of opioid and non-opioid substance use disorders and benzodiazepine co-medication. These patient characteristics did not vary as much between those with and without dysthymia or anhedonia. Overall, this suggests patients with depression and vital exhaustion have more risk factors for poor prescription opioid outcomes. This finding is consistent with the revised CDC opioid prescribing guideline which recommends screening for “depression and mental health conditions” prior to initiating long-term opioid therapy.43

The sampling strategy for Pathways was designed to generate a prospective cohort enriched with participants who are at elevated risk for LTOT and worsening or incident mood disorders by requiring 30-90 days of prescription opioid use at baseline. Based on sample characteristics at baseline, such as a 20.3% prevalence of past year depression and 66.1% prevalence of daily opioid use, the sampling approach worked and we will be able to test if prescription opioid use leads to incident depression when the 6-month and 12-month follow-up data are available.

Limitations:

Except for EHR derived variables, measures were based on self-report, which could be biased by decaying recall and social desirability. The length of the baseline survey precluded use of diagnostic interviews for some psychiatric conditions, including GAD and PTSD. Potential participants were recruited from two large midwestern health care systems and results may not generalize to other regions, particularly where opioid prescribing is very restrictive.

Conclusions:

To our knowledge, the Pathways cohort is the only on-going U.S. cohort study of noncancer pain patients, all of whom were beginning a new period of prescription opioid use at baseline. Participants in Pathways were all starting a new period of prescription opioid use lasting 30 to 90 days and the descriptive results indicate patients who use opioids for this period of time are highly complex with multiple psychiatric and substance use comorbidities, lower educational attainment and more disability and unemployment. As longitudinal data is added, the Pathways cohort will become a valuable resource for studies of the consequences of long-term, frequent prescription opioid use.

Supplementary Material

1

Table 4.

Pathways Study participants characteristics by past year dysthymia at baseline

Past Year Dysthymia (n=71) No Past Year Dysthymia (n=967) p-value SMD%
Demographics
Age (mean, SD) 53.5 (11.5) 55.0 (11.3) 0.2986 12.7
Gender 0.5087
  Man 46.4% (33) 42.4% (409) 8.1
  Woman 53.6% (38) 57.6% (556) 8.1
Race 0.9282
  White 75% (53) 75.1% (705) 0.2
  Black 23.3% (16) 22.5% (211) 1.9
  Other <5 2.4% (22) 4.8
Ethnicity 0.4947
  Hispanic or Latino <5 2.8% (27) 1.3
  Not Hispanic or Latino 97.4% (69) 97.2% (919) 1.3
Employment 0.2177
  Working now 23.9% (17) 25.9% (248) 4.6
  No work now 27.6% (19) 19.4% (186) 19.5
  Disability due to pain 24.5% (17) 19.6% (188) 11.8
  Disability other reason <5 11.9% (114) 19.4
  Retired 17.8% (13) 23.3% (224) 13.8
Education 0.4660
  No College 33.9% (24) 30.1% (291) 8.0
  Some College 40% (28) 47.6% (459) 15.3
  Bachelor or higher 26.1% (18) 22.3% (215) 9.0
Household income 0.1746
  <$20,000 39% (25) 31.1% (283) 16.5
  $20,000-$49,999 31.5% (20) 25.4% (231) 13.5
  $50,000-99999 19.5% (12) 25.6% (232) 14.5
  100K + 10% (6) 17.9% (163) 22.9
Marital Status 0.7386
  Married/live with partner 57.3% (40) 52.6% (503) 9.4
  Widow/Div/Sep 28.1% (20) 30.3% (290) 4.8
  Never married 14.6% (10) 17.1% (163) 6.8
Pain measures
High pain severity 31.5% (22) 28.6% (276) 0.6042 6.3
High pain interference 35% (25) 32.9% (317) 0.7051 4.6
BPI # pain sites—mean (SD) 7.2 (4.2) 5.9 (3.6) 0.0146 32.8
EHR pain measures
Arthritis 65.4% (46) 65.9% (637) 0.9406 0.9
Back pain 62.5% (44) 59.5% (575) 0.6159 6.2
Musculoskeletal pain 56.1% (40) 54.6% (528) 0.8086 3.0
 Neuropathies 19.0% (13) 15.7% (152) 0.4679 8.6
 Migraine/headache 21.2% (15) 13.3% (129) 0.0642 21.0
Fibromyalgia <5 6.5% (63) 0.8632 2.2
Chronic Pain 50.8% (36) 43.5% (421) 0.2324 14.7
Prescription opioid measures
Daily OAU 78.0% (54) 67.0% (632) 0.0585 24.9
Daily MME ≥50 mg 23% (15) 16.9% (144) 0.2104 15.2
High PODS score 30.2% (21) 16% (153) 0.0022 34.2
Non-opioid therapies (from survey)1 .471
  0 <5 8.2% (79) 15.4
  1 29.6% (21) 31.7% (306) 4.4
  2 37.8% (27) 38.6% (373) 1.5
  3 28.1% (20) 21.6% (209) 15.1
Non-opioid therapies from EHR2 0.5069
  0 13.2% (9) 14.9% (144) 4.9
  1 12.3% (9) 16.2% (156) 10.9
  2 19.0% (13) 21.9% (212) 7.3
  3 30.2% (21) 22.5% (217) 17.6
  4 15.4% (11) 16.8% (162) 3.7
  5 9.9% (7) 7.6% (74) 8.1
  6 <5 <5 6.0
Psychiatric conditions
Depression past year 47.5% (33) 18.4% (178) <.0001 65.3
Depression lifetime 63.9% (45) 29.4% (285) <.0001 73.6
Bipolar past year 11.8% (8) 1.6% (15) <.0001 41.9
Bipolar lifetime 15.2% (11) 4.2% (40) <.0001 38.0
Anhedonia 55.8% (39) 26.9% (255) <.0001 61.4
Vital Exhaustion 41.4% (29) 16.4% (158) <.0001 57.3
GAD positive 43.6% (31) 17.7% (170) <.0001 58.7
PTSD positive 29.1% (20) 14.7% (139) 0.0015 35.3
AUDIT-C positive 24.6% (16) 21.5% (193) 0.5599 7.3
Opioid use disorder past year <5 <5 0.0652 11.4
Opioid use disorder lifetime <5 3.4% (32) 0.5035 1.3
Any non-opioid SUD past year <5 0.9% (9) 0.0149 18.1
Any non-opioid SUD lifetime 18.2% (13) 9.5% (92) 0.0188 25.5
Current Smoker 32.1% (23) 26.2% (252) 0.2781 13.0
  Co-medications
Benzodiazepine 29.7% (21) 26.2% (253) 0.5147 7.9
Any antidepressant 63.4% (45) 46.1% (446) 0.0050 35.2
Any psychotherapy <5 2.7% (26) 0.0517 13.4
Psychosocial
High emotional support 39.2% (28) 54.7% (524) 0.0120 31.3
High Social functioning <5 12.5% (119) 0.0339 31.5
Sleep disorder (PSQI)—mean (SD) 11.8 (4.2) 10.0 (4.3) 0.0017 41.8
Comorbidity Index—mean (SD) 1.9 (2.2) 1.8 (2.3) 0.7882 3.4
Open in a new tab
1)

survey based measures of CBD use or cannabis use for pain, chiropractic, PT

2)

EHR based measures of chiropractic, interventional pain, NSAID, gabapentin, muscle relaxant, PT

Perspective:

This study reports baseline characteristics of a new prospective, non-cancer pain cohort study. Risk factors for adverse opioid outcomes were most common in those with depression and vital exhaustion and less common in dysthymia and anhedonia. Baseline data highlight the complexity of patients receiving long-term opioid therapy for non-cancer pain.

Highlights.

  • The Pathways Study is a new U.S. non-cancer pain cohort

  • Nearly two-thirds of participants are daily prescription opioid users

  • Fewer risk factors for adverse opioid outcomes in dysthymia and anhedonia

  • Risk factors for adverse opioid outcomes associated with depression and vital exhaustion

ACKNOWLEDGEMENTS

The authors thank study participants for their time and effort.

The authors thank the many student research assistants who contributed to recruiting participants and conducting phone surveys.

DISCLOSURES

This work was supported by National Institute on Drug Abuse grant R01DA043811

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of Interest: All authors report no conflicts of interest relevant to this research study

REFERENCES

  • 1.Scherrer JF, Svrakic DM, Freedland KE, et al. Prescription opioid analgesics increase the risk of depression. Journal of General Internal Medicine. 2014;29(3):491–499. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Scherrer JF, Salas J, Copeland LA, et al. Prescription opioid duration, dose, and increased risk of depression in 3 large patient populations. Annals of Family Medicine. 2016;14:54–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Rosoff DB, Smith GD, Lohoff FW. Prescription Opioid Use and Risk for Major Depressive Disorder and Anxiety and Stress-Related Disorders: A Multivariable Mendelian Randomization Analysis. JAMA Psychiatry. 2021;78(2):151–160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Appels A, Mulder P. Fatigue and heart disease. The association between vital exhaustion and past, present and future coronary artery disease. Journal of Psychosomatic Research. 1989;33:727–738. [DOI] [PubMed] [Google Scholar]
  • 5.Meesters C, Appels A. An interview to measure vital exhaustion II. reliability and validity of the interview and correlations of vital exhaustion with personality charateristics. Psychology and Health. 1996;11:573–581. [Google Scholar]
  • 6.Kornerup H, Marott JL, Schnohr P, Boysen G, Barefoot J, Prescott E. Vital exhaustion increases the risk of ischemic stroke in women but not in men: results from the Copenhagen City Heart Study. J Psychosom Res. 2010;68(2):131–137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Scherrer JF, Salas J, Lustman PJ, Burge S, Schneider FD, Investigators rRNoTR. Change in opioid dose and change in depression in a longitudinal primary care patient cohort. Pain 2015;156:348–355. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Von Korff M, Shortreed SM, LeResche L, et al. A longitudinal study of depression among middle-aged and senior patients initiating chronic opioid therapy. J Affect Disord. 2017;211:136–143. [DOI] [PubMed] [Google Scholar]
  • 9.Campbell G, Mattick R, Bruno R, et al. Cohort protocol paper: the Pain and Opioids In Treatment (POINT) study. BMC Pharmacol Toxicol. 2014;15:17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Campbell G, Nielsen S, Bruno R, et al. The Pain and Opioids IN Treatment study: characteristics of a cohort using opioids to manage chronic non-cancer pain. Pain. 2015;156(2):231–242. [DOI] [PubMed] [Google Scholar]
  • 11.Scherrer JF, Salas J, Miller-Matero LR, et al. Long-term prescription opioid users’ risk for new-onset depression increases with frequency of use. Pain. 2022;163(8):1581–1589. [DOI] [PubMed] [Google Scholar]
  • 12.Younger JW, Chu LF, D’Arcy N, Trott K, Jastrzab LE, Mackey SC. Prescription opioid analgesics rapidly change the human brain. Pain. 2011;152(8):1803–1810. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Upadhyay J, Maleki N, Potter J, et al. Alterations in brain structure and functional connectivity in prescription opioid-dependent patients. Brain. 2010;133:2098–2114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Morasco BJ, O’Hearn D, Turk DC, Dobscha SK. Associations between prescription opioid use and sleep impairment among veterans with chronic pain. Pain Med. 2014;15(11):1902–1910. [DOI] [PubMed] [Google Scholar]
  • 15.Filiatrault ML, Chauny JM, Daoust R, Roy MP, Denis R, Lavigne G. Medium Increased Risk for Central Sleep Apnea but Not Obstructive Sleep Apnea in Long-Term Opioid Users: A Systematic Review and Meta-Analysis. J Clin Sleep Med. 2016;12(4):617–625. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Lee M, Silverman SM, Hansen H, Patel VB, Manchikanti L. A comprehensive review of opioid-induced hyperalgesia. Pain Physician. 2011;14(2):145–161. [PubMed] [Google Scholar]
  • 17.Smith HS, Elliott JA. Opioid-induced androgen deficiency (OPIAD). Pain Physician. 2012;15:ES145–ES156. [PubMed] [Google Scholar]
  • 18.Scherrer JF, Ahmedani B, Autio K, et al. The Prescription Opioids and Depression Pathways Cohort Study. J Psychiatr Brain Sci. 2020;5:1–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Keller S, Bann CM, Dodd SL, Schein J, Mendoza TR, Cleeland CS. Validity of the brief pain inventory for use in documenting the outcomes of patients with noncancer pain. The Clinical journal of pain. 2004;20(5):309–318. [DOI] [PubMed] [Google Scholar]
  • 20.Cleeland CS. The Brief Pain Inventory User Guide. In. Houston, TX: University of Texas M.D. Anderson Cancer Center; 1991. [Google Scholar]
  • 21.Banta-Green CJ, Von Korff M, Sullivan MD, Merrill JO, Doyle SR, Saunders K. The prescribed opioids difficulties scale: a patient-centered assessment of problems and concerns. The Clinical journal of pain. 2010;26(6):489–497. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Bucholz KK, Cadoret R, Cloninger CR, et al. A new, semistructured psychiatric interview for use in genetic-linkage studies - A report on the reliability of the Ssaga. Journal of Studies on Alcohol. 1994;55:149–158. [DOI] [PubMed] [Google Scholar]
  • 23.Spitzer RL, Kroenke K, Williams JBW, Lowe B. A brief measure for assessing generalized anxiety disorder: The GAD-7. Arch Intern Med. 2006;166:1092–1097. [DOI] [PubMed] [Google Scholar]
  • 24.Prins A, Bovin MJ, Smolenski DJ, et al. The Primary Care PTSD Screen for DSM-5 (PC-PTSD-5): Development and Evaluation Within a Veteran Primary Care Sample. J Gen Intern Med. 2016;31(10):1206–1211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Snaith RP, Hamilton M, Morley S, Humayan A, Hargreaves D, Trigwell P. A scale for the assessment of hedonic tone the Snaith-Hamilton Pleasure Scale. Br J Psychiatry. 1995;167(1):99–103. [DOI] [PubMed] [Google Scholar]
  • 26.Franken IH, Rassin E, Muris P. The assessment of anhedonia in clinical and non-clinical populations: further validation of the Snaith-Hamilton Pleasure Scale (SHAPS). J Affect Disord. 2007;99(1-3):83–89. [DOI] [PubMed] [Google Scholar]
  • 27.Bush K, Kivlahan DR, McDonell MB, Fihn SD, Bradley KA. The AUDIT alcohol consumption questions (AUDIT-C): an effective brief screening test for problem drinking. Ambulatory Care Quality Improvement Project (ACQUIP). Alcohol Use Disorders Identification Test. Arch Intern Med. 1998;158(16):1789–1795. [DOI] [PubMed] [Google Scholar]
  • 28.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. . Journal of Chronic Diseases. 1987;40:373–383. [DOI] [PubMed] [Google Scholar]
  • 29.Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Medical care. 2005;43(11):1130–1139. [DOI] [PubMed] [Google Scholar]
  • 30.Sharabiani MT, Aylin P, Bottle A. Systematic review of comorbidity indices for administrative data. Medical care. 2012;50(12):1109–1118. [DOI] [PubMed] [Google Scholar]
  • 31.PROMIS SF v.2.0-Emotional Support 4a scale. http://www.healthmeasures.net/search-view-measures?task=Search.search. Accessed 2020 Apr 22.
  • 32.Hahn EA, Devellis RF, Bode RK, et al. Measuring social health in the patient-reported outcomes measurement information system (PROMIS): item bank development and testing. Qual Life Res. 2010;19(7):1035–1044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Germain A, Hall M, Krakow B, Shear MK, Buysse DJ. A brief Sleep Scale for Posttraumatic Stress Disorder: Pittsburgh Sleep Quality Index Addendum for PTSD. Anxiety Disorders. 2005;19:233–244. [DOI] [PubMed] [Google Scholar]
  • 34.Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika Trust. 1983;70:41–55. [Google Scholar]
  • 35.Curtis LH, Hammill BG, Eisenstein EL, Kramer JM, Anstrom KJ. Using inverse probability-weighted estimators in comparative effectiveness analysis with observational databases. Medical care. 2007;45:S103–S107. [DOI] [PubMed] [Google Scholar]
  • 36.Xu S, Ross C, Raebel MA, Shetterly S, Blanchette C, Smith D. Use of stabilized inverse propensity scores as weights to directly estimate relative risk and its confidence intervals. Value Health. 2010;13:273–277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Austin PC, Stuart EA. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Stat Med. 2015;34:3661–3679. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Scherrer JF, Salas J, Miller-Matero LR, et al. Long-term Prescription Opioid Users Risk for New Onset Depression Increases with Frequency of Use PAIN. 2021;In Press. [DOI] [PubMed] [Google Scholar]
  • 39.Sullivan MD. Why does depression promote long-term opioid use? Pain. 2016;157(11):2395–2396. [DOI] [PubMed] [Google Scholar]
  • 40.Sullivan MD EM, Zhang L, Unutzer J, Wells KB. Association between mental health disorders, problem drug use, and regular prescription opioid use. Archives of Internal Medicine. 2006;166:2087–2093. [DOI] [PubMed] [Google Scholar]
  • 41.Bertin C, Delage N, Rolland B, et al. Analgesic opioid use disorders in patients with chronic non-cancer pain: A holistic approach for tailored management. Neurosci Biobehav Rev. 2021;121:160–174. [DOI] [PubMed] [Google Scholar]
  • 42.Davis MA, Lin LA, Liu H, Sites BD. Prescription Opioid Use among Adults with Mental Health Disorders in the United States. J Am Board Fam Med. 2017;30(4):407–417. [DOI] [PubMed] [Google Scholar]
  • 43.Dowell D, Ragan KR, Jones CM, Baldwin GT, Chou R. Prescribing Opioids for Pain - The New CDC Clinical Practice Guideline. The New England journal of medicine. 2022;387(22):2011–2013. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

1
NIHMS1941170-supplement-1.docx (35KB, docx)

RESOURCES