A Retrospective Study of Psychotropic Drug Use and Prescription Opioid Initiation Among Older Adults

Abstract

Individuals with mental health problems may be more vulnerable to using prescription opioids than their counterparts. Therefore, the main objective of this study was to assess the initiation of prescription opioids in older adults who used psychotropic drugs compared with those who did not. The authors used a retrospective cohort design and included a sample of older adults enrolled in Pennsylvania's Pharmaceutical Assistance Contract for the Elderly program who did not use prescription opioids in 2013. Using pharmacy claims, patients who used anxiolytics/sedatives/hypnotics (n = 13,512) or antidepressants (n = 17,492) between October and December 2013 were identified and compared with those who did not use anxiolytics/sedatives/hypnotics (n = 114,091) or antidepressants (n = 110,111) during that period, to determine the incidence of prescription opioid use in 2014. Chi-square tests and multivariate logistic regressions were performed for analyses. Compared with patients who did not use anxiolytics/sedatives/hypnotics, those who used were more likely to initiate prescription opioids (15.0% versus 22.0%, P < .0001). Similarly, compared with patients who did not use antidepressants, those who used were more likely to initiate prescription opioids (14.7% versus 21.9%, P < .0001). Multivariate logistic regression indicated that the odds of prescription opioid initiation increased with anxiolytic/sedative/hypnotic use by 44% (AOR = 1.44; P < .0001) and antidepressant use by 48% (AOR = 1.48; P < .0001) among older adults after adjusting for potential confounding variables. Results showed that prescription opioid initiation is associated with prior anxiolytic/sedative/hypnotic or antidepressant use among older adults. Patients with mental health problems should also be queried about pain experiences for effective treatment.

Introduction

The opioid epidemic has received much attention in the United States. The national overdose deaths involving prescription opioids alone increased from 12,796 to 17,029 from 2007 to 2017.¹ After a continuous surge in opioid-related overdose deaths among Medicare Part D beneficiaries until 2017, a recent report has shown a decline from 2017 to 2019.² Although there was a drop in opioid-related overdose deaths, the number of beneficiaries receiving drugs for opioid use disorder and overdose treatment has continued to grow steadily in recent years,² making the opioid crisis still a complex challenge.

Prescription opioids are medications prescribed by health care providers to treat acute as well as chronic pain. Older adults have a higher prevalence of multiple chronic conditions and health care utilization compared with younger adults,³ making them more susceptible to be prescribed opioids. Common clinical indications for opioid treatment in older adults include musculoskeletal, trauma, postsurgical, and dental problems.⁴ More than a quarter of Medicare Part D beneficiaries received prescription opioids in 2019. Although the proportion of beneficiaries who received prescription opioids dropped from 31% to 26% from 2017 to 2019, a significant number received these opioids at high dosages. A total of 266,728 beneficiaries received high amounts of prescription opioids with a mean dosage of >120 morphine milligram equivalents per day for at least 90 days in 2019.² A previous cross-sectional study found that 20.7% of older adults enrolled in a state-funded program used prescription opioids in 2017, of whom 1.4% used opioids at high dosages.⁵ Among opioid-naive older adults, it has been found that once initiated, about 30% continued to use opioids requiring ≥2 prescriptions with ≥15 days' supply and 6% transitioned to chronic use (>90 days).⁶ Although it is generally safe and effective to use opioids at low doses and for short durations, if used at high dosages or for longer durations these medications have been associated with a number of adverse consequences such as respiratory depression, confusion, addiction,⁷ and overdose^8,9; among older adults, higher dosage opioid use also has been associated with falls and fractures.¹⁰ According to reports published by the Agency for Healthcare Research and Quality, overall opioid-related hospitalizations and emergency department visits increased significantly from 2010 to 2015 among adults aged 65 years and older.¹¹ Definitions of high-dose/chronic opioid use applied in different studies are provided in Table 1.

Table 1.

Definitions Used for High-Dose/Chronic Opioid Use in Different Studies

Study	Term	Definition	Participants
Office of Inspector General, 2020	High amounts of opioid use	Opioid use of an average morphine equivalent dose of >120mg a day for at least 3 months	Medicare Part D beneficiaries (65 years or older)
Khan et al, 2020	High-dosage opioid use	Opioid use of >90 morphine milligram equivalents (mme)/day for ≥90 consecutive days.	PACE/PACENET enrollees (65 years or older)
Oh et al, 2019	(1) Incident chronic use (2) Prevalent chronic use	(1) No opioid use at first visit, but initiated use during the study period and continued use during follow-up (2) Reported opioid use at baseline, and continued use during follow-up	National Alzheimer's Coordinating Center enrollees (65 years or older)
Musich et al, 2019	High-dose opioid use	Opioid use of >120 mme/day	AARP Medicare Supplement enrollees with AARP MedicareRX plans (65 years or older)
Musich et al, 2019	Chronic opioid use	≥2 prescriptions and >90 days' supply of opioids	AARP Medicare Supplement enrollees with AARP MedicareRX plans (65 years or older)
Saunders et al, 2010	Opioid dosage level	Classified into 4 levels: 0 mg/day, 1-<20 mg/day, 20-<50 mg/day, or 50+ mg/day	Group Health Cooperative enrollees (60 years or older)
Morasco et al, 2010	High-dose opioid use	Opioid use ≥180 mme per day for ≥90 consecutive days	Patients from Veterans Integrated Service Network-20

PACE, Pharmaceutical Assistance Contract for the Elderly; PACENET, PACE Needs Enhancement Tier.

Given the potential health outcomes associated with prescription opioid use, focusing on risk factors that influence such use becomes vital. One risk factor that has consistently been associated with opioid use is having mental health problems. Among adults, having a mental health disorder was more likely to be associated with prescription opioid use compared with those without such a disorder.^12
–14 Anxiety disorders and depression were found to be important risk factors for regular prescription opioid use.^15,16 Moreover, opioid users also were more likely than opioid non-users to use antidepressants^16
–18 and benzodiazepines.¹⁶ Similar associations have been found in older adults. Previous findings⁵ suggested an increased likelihood of prescription opioid use among anxiolytic/sedative/hypnotic users and antidepressant users compared with their counterparts among older adults. Another cross-sectional study found that among older adults, being depressed and taking benzodiazepines or sleep medications were important characteristics of high-dosage opioid users.¹⁹ Additionally, in a longitudinal study, Oh et al found that antidepressant use was correlated with incident chronic opioid use among older adults.¹⁸ Yet a reverse hypothesis also has been investigated, with a number of studies finding that chronic pain and prescription opioid use are significant predictors of mental health disorders.^20
–22 Moreover, as discussed in a review by Bair et al,²³ the association between pain and depression may be bidirectional, with shared biological pathways and neurotransmitters that suggest a complex relationship.²⁴ It is unclear if mental health issues precede pain or vice versa, or if both occur concurrently; however, the correlation between the two has been well documented. In addition to this correlation, prescribing of psychotropic medications and opioids among older adults is on the rise and has become concerning because of associated adverse health events. The American Geriatrics Society Beers Criteria, which list potentially inappropriate mediations for older adults, recommend cautious prescribing of psychotropic medications and opioids and central nervous system (CNS)-active medications.²⁵ The recommendations include avoiding CNS-active polypharmacy (combined use of ≥3 agents, from the following drug classes for >30 days consecutively: antidepressants, antiepileptics, antipsychotics, benzodiazepines, and opioids) that may increase the risk of falls and fractures in older adults, and avoiding concurrent use of opioids with benzodiazepines, which has been associated with increased risk of overdose.²⁵

Although most studies have focused on associations between mental health disorders and chronic or regular opioid use, little is known about the influence of psychotropic drug use on prescription opioid initiation among older adults. This study sought to assess the incidence of initiation of prescription opioids in older adults who used psychotropic drugs (anxiolytics/sedatives/hypnotics or antidepressants) compared with those who did not, using a retrospective cohort design. It is hypothesized that older adults who use anxiolytics/sedatives/hypnotics or antidepressants have a higher probability of initiating prescription opioids compared with those who do not use anxiolytics/sedatives/hypnotics or antidepressants.

Methods

Study population

This study used a retrospective cohort design and the sample included older adults who were continuously enrolled in Pennsylvania's Pharmaceutical Assistance Contract for the Elderly (PACE) from January 1, 2013, to December 31, 2015. PACE is a state-funded program administered by the Pennsylvania Department of Aging that provides assistance with paying for prescription medications to income-eligible Pennsylvania residents who are aged 65 years and older. Further information about the program is provided elsewhere.²⁶ All prescription claims for PACE enrollees that were submitted to PACE and Medicare Part D between January 1, 2013, and December 31, 2014, were included. Of the initial sample of 189,835 persons who were continuously enrolled in PACE between January 2013 and December 2014, older adults who had claims for prescription opioids in 2013 (n = 57,538) were excluded to ensure there was no prescription opioid use during the 1-year baseline period. Also excluded were those who had claims for antineoplastic agents (n = 4,655) to capture opioid initiators for non-cancer-related pain and users of buprenorphine or methadone (n = 39) because these medications are also used to treat opioid addiction. Following these exclusions, a final sample of 127,603 older adults was available for analysis. The study protocol was approved by the New England Independent Review Board (Needham, MA).

Measures

Data sets with pharmacy dispensing claims data, prescription drug attributes, PACE membership eligibility, and demographic data were linked to conduct this retrospective study.

Psychotropic drug use: Pharmacy dispensing claims data were used to identify the exposure variable, psychotropic drug use. Psychotropic drug classes were defined based on the American Hospital Formulary Service (AHFS).²⁷ AHFS classification numbers for anxiolytics/sedatives/hypnotics and antidepressants are 28:24 and 28:16.04, respectively. Using pharmacy dispensing claims, anxiolytic/sedative/hypnotic use was defined as having prescriptions filled for anxiolytics/sedatives/hypnotics (n = 13,512) between October 1 and December 31, 2013 and no anxiolytic/sedative/hypnotic use as having no prescriptions filled for anxiolytics/sedatives/hypnotics (n = 114,091) between October 1 and December 31, 2013. Similarly, antidepressant use was defined as having prescriptions filled for antidepressants (n = 17,492) between October 1 and December 31, 2013 and no antidepressant use as having no prescriptions filled for antidepressants (n = 110,111) between October 1 and December 31, 2013.

Prescription opioid initiation: Pharmacy dispensing claims data were used to identify the main outcome variable, prescription opioid initiation. Prescription opioid initiation was defined as having ≥1 prescriptions filled for opioids (nalbuphine, opium, oxymorphone, butorphanol, pentazocine, tapentadol, fentanyl patch, hydrocodone, meperidine, hydromorphone, codeine, levorphanol, morphine, oxycodone, tramadol, or dihydrocodeine) between January 1 and December 31, 2014. In contrast, prescription opioid non-initiators were those older adults who did not fill any prescriptions for opioids between January 1 and December 31, 2014. In addition, prescription opioid initiators were further grouped into 3 categories based on morphine equivalent conversion factors²⁸: use of opioids weaker than morphine (codeine, meperidine, pentazocine, tramadol, tapentadol), use of opioids equivalent to morphine (hydrocodone, nalbuphine, morphine, opium), and use of opioids stronger than morphine (butorphanol, fentanyl patch, hydromorphone, levorphanol, oxycodone, oxymorphone). Individuals who used opioids in >1 category were categorized into 3 mutually exclusive categories depending on the strongest opioid they used in 2014.

Potential confounding variables: Available information in the linked data set included age, sex, race, residence type, marital status, and PACE program subgroup – all of which were used as covariates in this study. Age was divided into 5 categories: 65 to 69 years, 70 to 74 years, 75 to 79 years, 80 to 84 years, or 85 years or older. Sex was dichotomized as male or female and race was classified as White, Black, or other race (American Indian or Alaska Native, Asian, Native Hawaiian or Pacific Islander, multiracial, other, or unknown). Residence type was categorized as living in one's own home, living in a rental house/an apartment, living in a nursing home/personal care home, living with a relative, or other/missing. Marital status was classified as single or widowed, married, or divorced/married, living separately. The PACE program subgroup was classified as either PACE or PACE Needs Enhancement Tier (PACENET); the subgroups differ in their benefit structure and income eligibility. Total annual income must be ≤$14,500 for a single person or ≤$17,700 for a married couple to be enrolled in PACE, whereas total annual income must be between $14,500 and $23,500 for a single person or between $17,700 and $31,500 for a married couple to be enrolled in PACENET, during the study period.²⁶

Statistical analyses

Distribution of demographic characteristics of the cohort was described by anxiolytic/sedative/hypnotic use status and antidepressant use status using chi-square tests. Distribution of prescription opioid initiation also was described by the main exposure variables and covariates. Further, multivariate logistic regression was conducted to identify factors associated with prescription opioid initiation. Multinomial logistic regression was performed to predict categories of prescription opioid use based on morphine equivalence by psychotropic drug use among opioid initiators. Adjusted odds ratios (AORs) with 95% confidence intervals (CIs) from the multivariate logistic regression were reported. All statistical analyses were performed using SAS version 9.4 (SAS Institute, Inc., Cary, NC).

Results

Among 127,603 PACE enrollees, 13,512 (10.6%) used anxiolytics/sedatives/hypnotics between October and December 2013. Compared with older adults who did not use anxiolytics/sedatives/hypnotics, those who used were more likely to be aged 80 years and older (P = .0012), female (P < .0001), White (P < .0001), single or widowed (P < .0001), and a PACE subgroup member (P = .0067) and less likely to own a home (P < .0001) (Table 2). Similarly, among all enrollees, 17,492 (13.7%) used antidepressants between October and December 2013. Compared with older adults who did not use antidepressants, those who used were more likely to be female (P < .0001), White (P < .0001), single or widowed (P < .0001), and a PACENET subgroup member (P = .0168), and less likely to be aged 75 and older (P < .0001) and own a home (P < .0001) (Table 2).

Table 2.

Descriptive Statistics by Anxiolytic Use and Antidepressant Use Status between October and December 2013 (n = 127,603)

Characteristics	No anxiolytic use (n = 114,091, 89.4%)	Anxiolytic use (n = 13,512, 10.6%)	P	No antidepressant use (n = 110,111, 86.3%)	Antidepressant use (n = 17,492, 13.7%)	P
Age (Years) 65–69 70–74 75–79 80–84 85 or older	9.0 19.2 22.5 22.5 26.8	8.1 18.9 22.4 23.2 27.5	0.0012	8.8 19.0 22.5 22.7 27.0	9.8 20.1 22.2 21.9 26.0	<.0001
Sex Female	69.6	83.0	<.0001	69.1	83.1	<.0001
Race/Ethnic origin White Black Other	83.0 7.1 9.9	89.6 2.5 7.9	<.0001	82.8 7.3 10.0	89.5 2.5 7.9	<.0001
Residence type Own Rent Nursing home or personal care home Live with relative Other/Missing	67.0 21.3 0.7 5.7 5.3	64.9 22.9 1.2 5.7 5.4	<.0001	67.5 21.1 0.5 5.6 5.3	62.3 23.7 2.1 6.3 5.6	<.0001
Marital status Single or widowed Married Divorced/Married, living separately	62.3 25.8 11.9	66.5 21.4 12.1	<.0001	62.3 26.0 11.7	65.9 21.0 13.1	<.0001
Program subgroup PACE (n = 47,820) PACENET (n = 79,783)	37.4 62.7	38.5 61.5	0.0067	37.6 62.4	36.7 63.3	0.0168

PACE, Pharmaceutical Assistance Contract for the Elderly; PACENET, PACE Needs Enhancement Tier.

For the entire cohort, 20,044 (15.7%) initiated prescription opioids in 2014. Bivariate analyses showed that among those who used anxiolytics/sedatives/hypnotics, 22.0% initiated prescription opioids and among those who did not use anxiolytics/sedatives/hypnotics, 15.0% initiated prescription opioids (P < .0001). Likewise, a significant difference in opioid initiation was observed between those who used antidepressants and those who did not (21.9% versus 14.7%, P < .0001). Among all age groups, opioid initiation incidence was lowest among persons aged 85 years and older. Females were more likely than males to initiate prescription opioids (16.4% versus 14.0%, P < .0001). Whites were more likely to initiate opioids compared with Blacks and other races (P < .0001). Those who were divorced or married but lived separately (16.4%, P = .0429) and those who were members of the PACENET subgroup (16.1%, P < .0001) were more likely to initiate opioids compared with their counterparts (Table 3).

Table 3.

Bivariate Associations of Demographic Characteristics and Psychotropic Drug Use with Prescription Opioid Initiation in 2014 (n = 127,603)

Characteristics	Total	Prescription opioid initiation (n = 20,044; 15.7%)	P
Age (Years) 65–69 70–74 75–79 80–84 85 or older	8.9 19.2 22.5 22.6 26.9	17.0 16.2 15.9 16.2 14.3	<.0001
Sex Female Male	71.0 29.0	16.4 14.0	<.0001
Race/Ethnic origin White Black Other	83.7 6.6 9.7	15.9 14.2 15.0	<.0001
Residence type Own Rent Nursing home or personal care home Live with relative Other/Missing	66.8 21.4 0.8 5.7 5.3	15.5 16.3 15.3 15.8 16.3	0.0210
Marital status Single or widowed Married Divorced/Married, living separately	62.8 25.3 11.9	15.6 15.8 16.4	0.0429
Program subgroup PACE PACENET	62.5 37.5	15.1 16.1	<.0001
Use of anxiolytics/sedatives/hypnotics between October and December 2013 Yes No	10.6 89.4	22.0 15.0	<.0001
Use of antidepressants between October and December 2013 Yes No	13.7 86.3	21.9 14.7	<.0001

PACE, Pharmaceutical Assistance Contract for the Elderly; PACENET, PACE Needs Enhancement Tier.

Multivariate logistic regression indicated that the odds of prescription opioid initiation increased with anxiolytic/sedative/hypnotic use by 44% (AOR = 1.44; P < .05) and antidepressant use by 48% (AOR = 1.48; P < .05) after adjusting for potential confounding variables (Table 4). Multinomial logistic regression showed that among opioid initiators, there was no significant association between anxiolytic/sedative/hypnotic use and prescription opioid initiation based on morphine equivalence. However, opioid initiators who used antidepressants were 11% (AOR = 1.11) more likely to use opioids equivalent to morphine and 13% (AOR = 1.13) more likely to use opioids stronger than morphine compared with those who did not use antidepressants (Table 5).

Table 4.

Multivariate Logistic Regression Predicting Prescription Opioid Initiation (n = 127,603)

Characteristics	AOR (95% CI)
Age (Years) 65–69 70–74 75–79^* 80–84^* 85 or older^*	1.00 0.94 (0.89, 1.00) 0.93 (0.87, 0.98) 0.94 (0.89, 0.99) 0.81 (0.77, 0.86)
Sex Female vs Male^*	1.18 (1.14, 1.23)
Race/Ethnic origin White Black Other^*	1.00 0.94 (0.88, 1.00) 0.94 (0.89, 0.99)
Residence type Own Rent^* Nursing home or personal care home Live with relative Other/Missing	1.00 1.06 (1.02, 1.10) 0.90 (0.75, 1.07) 1.03 (0.96, 1.10) 1.07 (1.00, 1.15)
Marital status Single or widowed^* Married Divorced/Married, living separately	0.96 (0.92, 0.99) 1.00 1.00 (0.95, 1.06)
Program subgroup PACE versus PACENET^*	0.92 (0.89, 0.95)
Use of anxiolytics/sedatives/hypnotics between October and December 2013 (Yes versus No)^*	1.44 (1.37, 1.50)
Use of antidepressants between October and December 2013 (Yes versus No)^*	1.48 (1.42, 1.54)

AOR, adjusted odds ratio; CI, confidence interval; PACE, Pharmaceutical Assistance Contract for the Elderly; PACENET, PACE Needs Enhancement Tier.

P < 0.05.

Table 5.

Multinomial Logistic Regression Predicting Categories of Prescription Opioid Initiation Based on Morphine Equivalence by Psychotropic Drug Use Among Prescription Opioid Initiators (n = 20,044)

Characteristics	AOR (95% CI)
Anxiolytic/sedative/hypnotic use (Yes vs No) Initiation of opioids weaker than morphine Initiation of opioids equivalent to morphine Initiation of opioids stronger than morphine	1.00 0.92 (0.84, 1.01) 0.98 (0.89, 1.09)
Antidepressant use (Yes vs No) Initiation of opioids weaker than morphine Initiation of opioids equivalent to morphine^* Initiation of opioids stronger than morphine^*	1.00 1.11 (1.02, 1.20) 1.13 (1.03, 1.24)

AOR, adjusted odds ratio, CI, confidence interval.

P < 0.05.

Discussion

This study examined the influence of anxiolytic/sedative/hypnotic use and antidepressant use on prescription opioid initiation among older adults using a retrospective cohort design. Results support the hypothesis that anxiolytic/sedative/hypnotic use and antidepressant use are associated with opioid initiation. Several studies have shown significant associations between depression or use of antidepressants and prescription opioid use among US adults, including 4 cross-sectional studies^12,15
–17 and 1 longitudinal study.¹³ Two studies also showed significant associations between anxiety and prescription opioid use.^15,16 Similarly, studies on older adults showed higher dose/chronic opioid users were more likely to be depressed or to take antidepressants and anxiolytics or sedatives.^5,18,19 The findings that anxiolytic/sedative/hypnotic use (AOR = 1.44, 95% CI: 1.37, 1.50) and antidepressant use (AOR = 1.48, 95% CI: 1.42, 1.54) between October and December 2013 significantly predicted prescription opioid initiation in 2014 support and extend the literature on this association. Older adults who took antidepressants also had a higher likelihood of using opioids equivalent to or stronger than morphine. Furthermore, 173 (0.86%) of the prescription opioid initiators in this study had chronic use during the study period (at least 90 consecutive days of opioid use from the day of initiation). The Institute of Medicine reported that older adults are prone to a number of chronic conditions such as hypertension, malignant neoplasms, diabetes, problems with joints, and heart disease.³ The proportion of older adults with anxiety disorder is estimated to be between 7% and 9.5%²⁹ and the proportion with major depression is between 1% and 5%,³⁰ which escalates substantially for those requiring home health care and older hospital patients.³⁰ In the co-occurrence of such psychiatric disorders, older adults with chronic conditions may experience high pain sensitivity and intensity,^31,32 which might explain the increased likelihood of opioid initiation among older patients with psychiatric disorders. This study did not evaluate pain using specific pain intensity or sensitivity measures. The inferences about pain were based exclusively on prescription opioid initiation.

In addition to use of anxiolytics/sedatives/hypnotics and antidepressants, several demographic variables were significantly associated with prescription opioid initiation. First, although anxiolytic/sedative/hypnotic users were more likely to be age 85 years and older compared with nonusers (27.5% vs 26.8%), this study found that there was a 19% (AOR = 0.81) lower probability of prescription opioid initiation among older adults in this age group versus the group aged 65 to 69 years. This finding is in line with Marra et al, who found that, compared with those aged 65 to 74 years, those who were 85 years or older were less likely to be administered prescription opioids in emergency departments and to be prescribed them at discharge.³³ Marra et al also showed decreased odds of opioid prescription administration among those aged 75 to 84 years. Likewise, the present study also showed decreased odds of opioid initiation among those aged 75 to 79 years (AOR = 0.93), 80 to 84 years (AOR = 0.94), and those who aged 85 years or older (AOR = 0.81) compared with those aged 65 to 69 years. These findings suggest that providers remain cautious about limiting opioids for older adults, likely reflecting awareness of and concerns about adverse outcomes.³³ Second, females were 18% more likely than males to initiate prescription opioids. Prior studies also have shown that females are more likely to use any pain medications than males^5,12,34; however, prior research also suggests that males are more likely to use prescription opioids at high dosages.^5,19 With respect to race, individuals of other races including American Indian or Alaska Native, Asian, Native Hawaiian or Pacific Islander, multiracial, other, or unknown (AOR = 0.94) were less likely to initiate prescription opioids compared with Whites, which was consistent with Davis et al's findings.¹² Compared with older adults living in self-owned homes, those living in rental homes had 6% higher odds of initiating prescription opioids. Moreover, members of the PACENET subgroup (who have comparatively higher income than the PACE subgroup) were at decreased odds of opioid initiation. Findings from a previous cross-sectional study also had demonstrated higher opioid use among older adults living in rental housing and among PACE members compared with their counterparts.⁵

There are some limitations in this study. First, the definitions of anxiolytic/sedative/hypnotic use, antidepressant use, and prescription opioid initiation were based on pharmacy dispensing claims data and there was no indication of whether patients actually consumed the medications as directed. Second, the study findings may not be generalizable to other populations because, compared with all US older adults, PACE enrollees are older and poorer, on average, and are more likely to be female. Third, although this study adjusted for a number of confounding variables, it did not control for factors such as health status and use of alcohol and other drugs that have been found to be associated with mental health disorders^35,36 as well as prescription opioid misuse,^13,37 as no data were available on these factors. Fourth, the exposure period for this study was defined as October through December 2013, and the impact of psychotropic medication use outside of this time period on opioid initiation was not evaluated. Further, prior studies have shown the possible effects of seasonal variation on human mood and social behavior,^38,39 which may affect psychotropic drug use. The exposure period did not capture psychotropic drug use across all seasons and may have caused non-differential misclassification bias on study findings. Moreover, to study opioid initiation, individuals known to have filled opioid prescriptions during a 1-year baseline period in 2013 were excluded from this study. However, it is possible that some individuals who met this study criterion may have used opioids prior to 2013.

Regardless of these limitations, this study has several strengths. A retrospective cohort design was used to evaluate the temporal sequence between the exposure and outcome variables, which has strengthened the findings. This study also used a large and diverse sample of older adults within Pennsylvania and adjusted for a wide variety of important confounding factors. Moreover, this study captured multiple anxiolytics, sedatives, hypnotics, antidepressants, and prescription opioids that are found in the market through the use of point-of-sale pharmacy claims data. Further, in addition to all prescriptions billed to PACE, prescriptions for PACE members that were billed to Medicare Part D but not PACE also were included to cover all prescription claims during the study period.

Conclusion

Study findings show that prescription opioid initiation is associated with prior use of anxiolytics/sedatives/hypnotics or antidepressants among older adults. Because anxiety and depressive disorders are common in the older population, there is a high risk that these patients are also experiencing chronic pain. Therefore, patients with mental health problems also should be queried about pain experiences for effective treatment. Moreover, for the best possible health care, older adults should be screened for psychiatric disorders and treated early if diagnosed so that the likelihood of opioid initiation decreases. This screening would ensure safer opioid use, prevent adverse outcomes associated with prescription opioid use/misuse, and ultimately promote positive aging.

Footnotes

Authors' Contributions

All authors have significantly contributed to the publication. Drs. Khan and Heller designed the study and the main conceptual ideas. Dr. Khan performed statistical analyses, interpreted the results, and wrote the manuscript. Dr. Heller supervised the data analyses and revised the manuscript critically for substantial intellectual content. Mr. Latty, Ms. LaSure, and Ms. Brown provided critical feedback and helped shape the manuscript. All authors approved the final version for publication.

Acknowledgments

We thank the Pennsylvania Department of Aging for its support in providing data for this study.

Author Disclosure Statement

The authors declare that there are no conflicts of interest.

Funding Information

No funding was received for this article.

References

Centers for Disease Control and

Prevention

, National Center for Health

Statistics

. Multiple cause of death 1999–2017. 2018. https://wonder.cdc.gov Accessed September 18, 2019 .

Office of Inspector

General

. Opioid use in Medicare Part

D continued to decline in

, but vigilance is needed as covid-19 raises new concerns. 2020. https://oig.hhs.gov/oei/reports/OEI-02-20-00320.pdf Accessed October 16, 2020 .

Institute of Medicine (US) Committee on the Future Health Care Workforce for Older Americans. Retooling for an aging America: Building the health care workforce. Washington, DC: National Academies Press (US), 2008.

Pasricha

, Tadrous

, Khuu

, et al. Clinical indications associated with opioid initiation for pain management in Ontario, Canada: a population-based cohort study. Pain, 2018; 159:1562–1568.

Khan

, Heller

, Latty

, Cadieux

, LaSure

, Brown

. Association between psychotropic drug use and prescription opioid use among older adults. Geriatr Nurs, 2020; S0197-4572(20):30138-5.

Musich

, Wang

, Slindee

, Kraemer

, Yeh

. Characteristics associated with transition from opioid initiation to chronic opioid use among opioid-naive older adults. Geriatr Nurs, 2019; 40:190–196.

Chau

, Walker

, Pai

, Cho

. Opiates and elderly: use and side effects. Clin Interv Aging, 2008; 3:273–278.

Rudd

, Aleshire

, Zibbell

, Gladden

. Increases in drug and opioid overdose deaths—United States, 2000–2014. Am J Transplant, 2016; 64:1378–1382.

Guerriero

. Guidance on opioids prescribing for the management of persistent non-cancer pain in older adults. World J Clin Cases, 2017; 5:73–81.

10.

Saunders

, Dunn

, Merrill

, et al. Relationship of opioid use and dosage levels to fractures in older chronic pain patients. J Gen Intern Med, 2010; 25:310–315.

11.

Weiss

, Heslin

, Barrett

, Izar

, Bierman

. Opioid-related inpatient stays and emergency department visits among patients aged 65 years and older, 2010 and 2015: statistical brief #244. Healthcare Cost and Utilization Project (HCUP) Statistical Briefs. Rockville, MD: Agency for Healthcare Research and Quality (US), 2018.

12.

Davis

, Lin

, Liu

, Sites

. Prescription opioid use among adults with mental health disorders in the United States. J Am Board Fam Med, 2017; 30:407–417.

13.

Sullivan

, Edlund

, Zhang

, Unutzer

, Wells

. Association between mental health disorders, problem drug use, and regular prescription opioid use. Arch Intern Med, 2006; 166:2087–2093.

14.

Morasco

, Duckart

, Carr

, Deyo

, Dobscha

. Clinical characteristics of veterans prescribed high doses of opioid medications for chronic non-cancer pain. Pain, 2010; 151:625–632.

15.

Sullivan

, Edlund

, Steffick

, Unutzer

. Regular use of prescribed opioids: association with common psychiatric disorders. Pain, 2005; 119:95–103.

16.

Frenk

, Lukacs

, Gu

. Factors associated with prescription opioid analgesic use in the US population, 2011–2014. Pain Med, 2018; 20:1338–1346.

17.

Parsells

, Cook

, Kaufman

, Anderson

, Rosenberg

, Mitchell

. Prevalence and characteristics of opioid use in the US adult population. Pain, 2008; 138:507–513.

18.

, Abner

, Fardo

, Freeman

, Moga

. Patterns and predictors of chronic opioid use in older adults: a retrospective cohort study. PLoS One, 2019; 14:e0210341.

19.

Musich

, Wang

, Slindee

, Kraemer

, Yeh

. Prevalence and characteristics associated with high dose opioid users among older adults. Geriatr Nurs, 2019; 40:31–36.

20.

Burke

, Mathias

, Denson

. Psychological functioning of people living with chronic pain: a meta-analytic review. Br J Clin Psychol, 2015; 54:345–360.

21.

Ohayon

, Schatzberg

. Using chronic pain to predict depressive morbidity in the general population. Arch Gen Psychiatry, 2003; 60:39–47.

22.

Scherrer

, Svrakic

, Freedland

, et al. Prescription opioid analgesics increase the risk of depression. J Gen Intern Med, 2014; 29:491–499.

23.

Bair

, Robinson

, Katon

, Kroenke

. Depression and pain comorbidity: a literature review. Arch Intern Med, 2003; 163:2433–2445.

24.

Feingold

, Brill

, Goor-Aryeh

, Delayahu

, Lev-Ran

. The association between severity of depression and prescription opioid misuse among chronic pain patients with and without anxiety: a cross-sectional study. J Affect Disord, 2018; 235:293–302.

25.

The 2019 American Geriatrics Society Beers Criteria^® Update Expert Panel. American Geriatrics Society 2019 updated AGS Beers criteria^® for potentially inappropriate medication use in older adults. J Am Geriatr Soc, 2019; 67:674–694.

26.

Pennsylvania Department of Aging. Annual report to the Pennsylvania General Assembly January 1–December , 2019. 2020. https://www.aging.pa.gov/publications/annualreports/Documents/2019%20PACE%20Annual%20Report.pdf Accessed August 2, 2021 .

27.

American Hospital Formulary Service. AHFS clinical drug information. 2020. Available from: https://www.ahfsdruginformation.com/ahfs-classification-drug-assignments/ Accessed August 2, 2021 .

28.

Centers for Medicare & Medicaid Services. Opioid morphine equivalent conversion factors. https://www.cms.gov/medicare/prescription-drug-coverage/prescriptiondrugcovcontra/downloads/opioid-morphine-eq-conversion-factors-march-2015.pdf Accessed April 20, 2020 .

29.

Reynolds

, Pietrzak

, El-Gabalawy

, Mackenzie

, Sareen

. Prevalence of psychiatric disorders in U.S. older adults: findings from a nationally representative survey. World Psychiatry, 2015; 14:74–81.

30.

Centers for Disease Control and Prevention. 2017. https://www.cdc.gov/aging/mentalhealth/depression.htm Accessed 9 August, 2018.

31.

Hermesdorf

, Berger

, Baune

, Wellmann

, Ruscheweyh

, Wersching

. Pain sensitivity in patients with major depression: differential effect of pain sensitivity measures, somatic cofactors, and disease characteristics. J Pain, 2016; 17:606–616.

32.

Katon

, Sullivan

, Walker

. Medical symptoms without identified pathology: relationship to psychiatric disorders, childhood and adult trauma, and personality traits. Ann Intern Med, 2001; 134(9 Pt 2):917–925.

33.

Marra

, Mazer-Amirshahi

, Mullins

, Pines

. Opioid administration and prescribing in older adults in U.S. emergency departments (2005–2015). West J Emerg Med, 2018; 19:678–688.

34.

Serdarevic

, Striley

, Cottler

. Sex differences in prescription opioid use. Curr Opin Psychiatry, 2017; 30:238–246.

35.

Choi

, DiNitto

, Marti

, Choi

. Relationship between marijuana and other illicit drug use and depression/suicidal thoughts among late middle-aged and older adults. Int Psychogeriatr, 2016; 28:577–589.

36.

Mackenzie

, Reynolds

, Chou

, Pagura

, Sareen

. Prevalence and correlates of generalized anxiety disorder in a national sample of older adults. Am J Geriatr Psychiatry, 2011; 19:305–315.

37.

Becker

, Sullivan

, Tetrault

, Desai

, Fiellin

. Non-medical use, abuse and dependence on prescription opioids among U.S. adults: psychiatric, medical and substance use correlates. Drug Alcohol Depend, 2008; 94:38–47.

38.

Harmatz

, Well

, Overtree

, Kawamura

, Rosal

, Ockene

. Seasonal variation of depression and other moods: a longitudinal approach. J Biol Rhythms, 2000; 15:344–350.

39.

Sun

, Tang

, Liljenback

, et al. Seasonal variation in the brain mu-opioid receptor availability. J Neurosci, 2021; 41:1265–1273.