Disparities in healthcare exist across race and sex in the United States, including the prescribing of opioids and benzodiazepines (BZDs). Women are more likely to be prescribed and fill BZD prescriptions than men, and non-White patients are less likely to be prescribed and fill BZDs compared to White patients.^1–4 Men of all races and Black patients are less likely to be prescribed and fill opioids,^4–6 and co-prescribing of opioid and BZD is more likely among women and less likely among Black patients.^7,8

Despite concerns around the use of BZD and opioids among older adults due to increased risk of adverse effects, research suggests that the use of these medications is increasingly prevalent in older adults.^9,10 While opioids may be considered appropriate for chronic pain treatment, BZDs, often used in the treatment of insomnia and anxiety, are rarely appropriate for long-term use in older adults. Despite the risks of concurrent use, the proportion of older adults receiving both has more than doubled in the last 15 years.⁷

Pharmacoequity, ensuring that all individuals have access to the appropriate medications required to manage their health needs, regardless of race, sex, age, and other socioeconomic factors, is of great importance.¹¹ Though literature suggests racial and sex disparities in the prescribing of opioids and BZDs, the effects of race and sex on the prescribing of these medications among older adults specifically remain unknown. Prior research has focused on the broader adult population, with little exploration of whether such patterns hold true among older adults, who have a higher prevalence of co-morbid conditions for which opioids and BZDs may be prescribed, yet are at heightened risk for harms. Furthermore, inadequate pain relief for older adults, particularly Black and Hispanic men, have both short- and long-term implications for pain and physical function.¹² Inadequate prescribing of these medications can limit participation in daily activities and result in other health issues. Recognizing that both underuse and overuse can cause harm in older adults, pharmacoequity maintains that neither skin color nor age or other socioeconomic factors should influence the provision of these medications. The objectives of this study were to evaluate potential disparities in prescribing of opioids and BZDs among older adults and to identify factors associated with co-prescribing of opioids and BZDs.

We identified 25,120 older adults for this study. (Table 1). Most were under 80 years old (65–69: 29.6%; 70–79: 47.7%; 80–89: 19.5%; 90+: 3.2%), women (61.3%), White (81.9%), and were seen in clinics located in rural counties (59.2%). Fifteen percent of older patients self-reported a fall in the last year.

Among older patients, 3.3% (n=833) had chronic opioid prescriptions, with 20% receiving multiple opioid prescriptions concurrently (Table 1). The median daily MMEs was 13.2, with 12.5% prescribed greater than 50 MMEs per day. The most prescribed opioids were tramadol (36.5%), hydrocodone (28.3%), and oxycodone (26.1%). Chronic opioid users (Figure 1, Table S1) were more likely to be White women (56.2% vs. 49.1%, p<0.001) and less likely to be Black men (3.2% vs. 4.9%, p=0.03), other race women (1.1% vs. 2.3%, p=0.02), or other race men (0.4% vs. 1.43%, p=0.01).

Among older patients, 3.8% (n=959) had chronic BZD prescriptions, with 7.9% receiving multiple BZD prescriptions (Table 1). The median daily DMEs was 7.6, with 25.4% being prescribed greater than 10 DMEs per day. The most prescribed BZDs were alprazolam (34.5%), clonazepam (23.3%), and lorazepam (22.8%). Chronic BZD users (Figure 1, Table S1) were more likely to be White women (68.1% vs. 48.6%, p<0.001) and less likely to be White men (24.3% vs. 32.9%, p<0.001), Black women (4.1% vs. 9.9%, p<0.001), or Black men (1.4% vs. 4.9%, p<0.001).

Among chronic opioid and BZD users, 15.4% and 13.4%, respectively were co-prescribed both medications. Select factors associated with co-prescribing in adjusted logistic regression models are presented in Figure 2. Full results are presented in Table S2 and S3. Few demographic factors were associated with co-prescribing. However, the directionality of association was relatively consistent across most factors.

Among chronic opioid users, Black race was associated with reduced odds being prescribed BZDs in unadjusted analyses (OR=0.42 [0.20–0.89]) along with prior fall history (OR=1.50 [1.01, 2.23], but both were attenuated in adjusted analyses. Anxiety, major depression, PTSD, and insomnia were associated with co-prescribing of BZDs in unadjusted and adjusted analyses.

Among chronic BZD users, having a primary care clinic in an urban county was associated with reduced odds of being prescribed opioids in both unadjusted and adjusted analyses (aOR=0.25 [0.13,0.50]). Medicaid use was associated with increased odds of being prescribed opioids in unadjusted analyses [OR=2.40 [1.45, 3.97], but this was attenuated in adjusted analyses. Among chronic BZD users, individuals who were co-prescribed opioids were nearly two times as likely to have fallen in the year prior (aOR=1.90 [1.23, 2.95]). Chronic pain was associated with an increased likelihood of opioid prescribing, along with co-morbid depression and PTSD, in unadjusted and adjusted analyses.

Consistent with other studies, our study showed significant differences in the use and prescribing of opioids and BZDs by sex, race, and geographic location.^1–8,14,15 Disparities in prescribing, are of great importance considering that historically marginalized individuals disproportionally experience worse health outcomes and suboptimal care.^11,16,17 Considering the potential for abuse and misuse, it is also critical to recognize that such disparities may also contribute to overprescribing.^18,19

Race and sex have long played a role in the prescribing of opioids and BZDs. Multiple studies have shown that Black/African American adults are less likely than their counterparts to receive opioids for pain management, even for severe pain.²⁰ Differences among race and sex are also shared among BZDs prescriptions. Several studies in the U.S. have consistently shown that White women are the most likely to receive BZDs.^1–4 Unlike these findings, one study in Europe revealed no significant differences in BZD use in older adults based on sex.²¹ However, these results may not be generalizable due to differences in social constructs between Europe and the U.S.

Our study adds to this literature by demonstrating that similar disparities in prescribing of opioids and BZDs exist in the older adult population, who have a higher prevalence of chronic pain and neuropsychiatric conditions. However, co-prescribing of both opioids and BZDs was more driven by co-morbid conditions, with little evidence of disparities by race and sex. While this may represent ongoing disparities in access to treatment and medications for non-Whites, these disparities may have inadvertently shielded them from the risks associated with BZD and prescription opioid overuse.^4,22 Benefits stemming from healthcare disparities are an atypical outcome but should not overshadow the underlying contributing factors.

Another potential factor contributing to the observed disparities in BZD use in older adults is the historical marketing of tranquilizers to White women in the late 1950s.^23,24 For example, a 1973 study investigating national patterns of psychotherapeutic drug use showed that women used minor tranquilizers – such as diazepam, chlordiazepoxide, and meprobamate – more than twice as often as men.²³ Historically, it has also been more culturally acceptable for women to report “psychic distress” and seek help, while men have tended to use other substances, such as alcohol and marijuana, to manage anxiety.^23,24 Furthermore, as these drugs entered the market, non-Whites and individuals with lower incomes did not have equal access and were not viewed as sympathetically by healthcare providers, further contributing to disparities over time.²⁴

Several limitations should be acknowledged. Most patients were seen in clinics located in rural counties, and thus we were unable to fully evaluate geographic disparities. Residency, specifically living in rural spaces, is a stronger predictor of having an opioid prescription, despite race.¹⁴ However, this geographic distribution of rurality is relatively representative of NC, with most counties being rural. Regardless, further research is warranted to corroborate this phenomenon, recognizing that these findings are complex to interpret.¹⁴ We were also unable to evaluate the influence of individual patient characteristics beyond those listed due to data restrictions, which only allowed us to access to clinic-level summary data for the overall 65+ population. We were also unable to evaluate influences on prescribing based on clinician- or clinic-level characteristics. However, the main strength of our study is that by using data from electronic health records, our findings highlight real-world patterns of medication prescribing for older adults seen in primary care.

The findings presented in our study combined with the findings of prior research demonstrate that disparities in the prescribing of opioids and BZDs also affect older adults. Whether a result of racism, bias, discrimination, or other prejudiced factors, these findings underscore the need for reform, especially given the higher prevalence of relevant co-morbid conditions in this population and research confirming that there is no difference in the prevalence of pain or other psychiatric symptoms across racial groups, and clinical measures may fail to integrate the differing experiences related to racial weathering.^25,26 Interventions to address these disparities might include training on equitable symptom assessment across patient groups or educational materials for discussing symptom management with input from a diverse group of stakeholders, that is inclusive of older patients and caregivers. Further research is needed to better understand prescriber perspectives on opioid and BZD use across patient groups to identify priority areas for such interventions.