Comparison of the Impacts of Under-Treated Pain and Opioid Pain Medication on Cognitive Impairment

Purpose: To guide clinicians in balancing the risks and benefits of opioids when treating pain, we conducted two systematic reviews: 1) the impact of pain on cognitive function, and 2) the impact of opioids on cognitive function. Methods: Part one addressed the impact of pain on cognitive impairment; Part two considered the impact of opioids on cognitive impairment. PubMed was used to search for eligible articles. For part one, 1786 articles were identified, of which 23 met our eligibility criteria. For part two, among 584 articles, 18 were found eligible. Results: For part one, 16 studies concluded that patients with chronic pain showed impaired cognitive function; six studies found that chronic pain does not worsen cognitive function; one study concluded that the impact of pain on cognitive function differs based on the underlying cognitive status. For part two, 15 studies found that using opioids to control pain did not cause significant cognitive impairment, while three studies concluded the opposite. Studies evaluating older subjects did not observe different results from those in the whole population for both reviews. Conclusion: The published literature indicates that moderate to severe pain can impair cognitive function, and that careful use of opioid analgesics in such subjects does not necessarily worsen cognition. Although our results are insufficient to support clear guidance due to heterogeneity of cohorts and outcomes, this study may assist primary care providers by rendering explicitly the factors to be considered by providers caring for this population with pain when opioids are considered.


Introduction And Background
The opioid crisis is a nationwide concern. Misused opioids can lead to overuse, substance use disorder, and serious health outcomes. However, the use of opioids is frequently considered when other treatment modalities fail to manage pain [1].
One of the main concerns related to opioids is cognitive impairment [2], especially, for the frail elderly population [3]. Ironically, undertreated pain itself may increase the risk of cognitive impairment [4].
Currently, recommendations to guide the judicious use of opioids in this context are limited. Beers Criteria [5] (potentially inappropriate medication use guideline from the American Geriatric Society) state that opioid analgesics can cause ataxia, impaired psychomotor function, syncope, and additional falls. It recommends providers to avoid using opioids except for pain management in the setting of severe acute pain, such as recent fractures or joint replacement. As this is a rather general statement, it is desirable to have a more practical, evidence-based recommendation.
Previous literature focuses on either the effect of opioids on cognitive impairment or the effect of undertreatment of pain on cognitive impairment, respectively. A comprehensive review of risk assessment on both aspects is conducted in the present systematic review.

Review Methods
This review project consists of two parts. Part one is a systematic review of the impact of pain on cognitive function. For this part of the review, the study group was identified as patients with pain and the control group was participants with no pain. Part two is a systematic review of the impact of opioids on cognitive function. For part two, the study group was patients being treated opioids alone or in combination with other analgesic drugs whereas the control group was people with no treatment or analgesics other than opioids.
For both parts, the principal outcome of interest was cognitive impairment.

Inclusion criteria
Age inclusion criteria were set for age 19 or older, with no upper limit. Subgroup analysis for the age group over 65 years old was performed. All opioid formulations, including oral, intravenous, sublingual, and transdermal, were included in this review.

Exclusion criteria
Articles based on self-reported cognitive impairment were excluded. Cancer pain or pain in palliative care was not considered in our review. Articles focusing on the outcome of neuropsychological conditions, such as delirium, agitation, were excluded, as we focused on long-term impairment of cognitive function, rather than transient or short-term complications. Review, systematic review, or meta-analysis articles were excluded. Studies not available in English full text were also excluded.

Search strategy
We used Pubmed to search for studies published up to December 2020, as it appears in Table 1. No difference in retrieved articles was found using Medline search.

Pain and Cognitive Impairment (Part One)
We found 1786 articles using the Pubmed search engine, as it appears in Figure 1. Forty-two were identified as being relevant to the topic. Nineteen of the 40 articles were excluded for the following reasons. Eight articles used self-reported cognitive impairment. One focused on the study population younger than 19 years old. One studied dissociative symptom. Two were excluded due to study design, being a case series and a systematic review. One study was excluded due to high likelihood of bias, as it evaluated the effect of whiplash associated disorder (WAD) on driving function, which is more likely related to the physical mobility limitation of whiplash injury than pain. Four studies were excluded as they focused on neuropsychological conditions such as delirium. Two were excluded due to lack of the full text. Some 23 articles  were included in the final review.

Opioids and Cognitive Impairment (Part Two)
We found 584 articles in our Pubmed search, as it appears in Figure 2. After screening for the relevance to the topic, 37 articles remained. Among those, 19 articles were excluded for the following reasons. Eight studies were review articles, one was a letter to the editor, and one was a case report. Five articles were not available with English full text. Four studies did not have any full text. One study was excluded as it considered the outcome of hallucination. Some 18 articles [29][30][31][32][33][34][35][36][37][38][39][40][41][42][43][44][45][46] were included in the final review.

Demographic information
Demographic information is described in Table 2. Out of the 23 studies being reviewed, eight studied the geriatric population over 65 years old. Three studies did not have control group, and three studies had multiple study subgroups. In general, more females were studied than males.

Study design
Fourteen out of 23 studies reviewed were cross-sectional studies. The remainder were four prospective cohort studies, three case-control studies, and two retrospective cohort studies.

Pain type, chronicity, and intensity
Musculoskeletal pain was the most studied. Nine articles studied musculoskeletal pain such as neck, back, or mixed pain. Three studied WAD. Two studied mixed type of pain, one focused on chronic fatigue syndrome (CFS), and one focused on somatoform pain disorder (SPD). Eight studies did not report the type of pain included.
Most of the reviewed articles studied chronic pain except for one article that included patients with acute or chronic pain.
Moderate intensity pain was most commonly studied. Eight articles included patients with moderate pain. Three studied moderate to severe pain, two studied mild pain, and one focused on severe pain. Six articles did not report pain intensity.

Cognitive testing
Various types of cognitive testing modalities were used in the reviewed studies, as it appears in

Results
Sixteen out of 23 reviewed studies concluded that patients with chronic pain have worse cognitive function than patients without pain. Six studies presented opposite findings, i.e. that chronic pain does not worsen cognitive function. Also, one study [18] stated that the impact of pain on cognitive function differs based on the underlying cognitive impairment. In this study, the Alzheimer's group showed a significant correlation between pain and cognitive decline; however, the cognitively intact group did not show a significant correlation between those two.
The studies concluding that pain does not impair cognitive function enrolled a relatively small number of participants, as it appears in Table 3. Confounding factors such as underlying cognitive impairment, malingering, and level of education or intelligence were not adequately addressed. The importance of such confounding factors is attested to one study [18], which concluded that the relationship between pain and cognitive impairment differs based on the underlying cognitive function level. Also, one study [10] had a short follow-up time, which makes it hard to predict long-term effects. Regarding study design, they were mostly (61%) cross-sectional studies.  Age was not a factor influencing these different results. Four out of those seven studies, with the conclusion that pain does not worsen cognitive function, studied the geriatric population.

Studies focusing on ages over 65 years old
As described in Table 4, all eight articles studying the geriatric population focused on chronic pain, and musculoskeletal pain was studied in three of these eight. Seven articles were cross-sectional studies. Four articles concluded that chronic pain worsens cognitive function, whereas the others concluded the opposite that chronic pain does not worsen cognitive function. However, two studies with large sample sizes [25,27] both found that chronic pain is related to cognitive impairment. Those articles reporting that pain does not worsen cognitive function had limitations in study design with small sample sizes, and other confounding factors.

Studies about whiplash associated disorder related pain
Three studies focused on WAD-related pain, as it appears in Table 5. Two [7,20] out of those three studies concluded that WAD-related pain worsens cognitive function. One study [10] reported that WAD-related pain is not associated with cognitive function but did not assess the long-term effects.

Comparison of the results with other systematic reviews
A systematic review by de Aguiar et al. [47] concluded that persistent pain was not associated with cognitive impairment in geriatric populations. However, persistent pain was associated with cognitive decline in cases of follow-up length less than 4.5 years. Differences in inclusion and exclusion criteria exist between this systematic review and ours. In our review, transient changes or subjectively reported cognitive function were excluded. Headache was not included in our study due to its often intermittent nature.

Demographic information
Demographic information is available in Table 6. Out of 18 reviewed studies, four studied geriatric populations over 65 years old. Five studies did not have a control group, but simply followed a cohort of patients across time. There was no significant male or female preponderance in the patient population.

Study design
Prospective cohort studies (14 out of 18 studies being reviewed) were most common, followed by two crosssectional studies, one retrospective cohort study, and one randomized controlled trial.

Pain type, chronicity, and intensity
All reviewed studies focused on chronic pain unless otherwise specified. Mixed pain etiologies (musculoskeletal, visceral, and neuropathic) were most studied in nine articles. Two studies included patients with chronic low back pain (cLBP). Six studies did not report the type of pain. One study included only participants without any pain. Study subjects had moderate intensity of pain in 11 studies, and mild pain in two studies. The remaining five studies did not report pain intensity.

Opioids
Diverse opioid analgesics were employed in seven out of 18 reviewed studies. There was no consistency in the type of opioids, which caused difficulty in merging their findings. Oxycodone, morphine, transdermal fentanyl, and transdermal buprenorphine were used by subjects in two studies each. Codeine was used in one study, and there were two studies that did not present information about the type of opioid analgesics. Opioid doses also differed widely between studies. In most studies, there was very limited information about the duration of treatment and the other concomitantly used pain medications. Figure 4 shows what type of cognitive tests were addressed in the studies. The driving function assessment test was most used to assess cognitive function, followed by the trail-making test and the reaction time test. Five articles assessed driving function. Three assessed driving function indirectly, with the German national recommendations test battery, which is a panel of three psychophysical tests consisting of an attention cognitive test, a determination test, and a tachistoscopic perception test. The tests that were commonly used in the articles about the impact of pain on cognitive impairment were used much less in this part of the review, such as the Stroop task. This lack of consistency limits the ability to aggregate the results.

Results
Fifteen out of 18 articles concluded that using opioids does not cause cognitive impairment. Two articles concluded that opioids do worsen cognitive function. One study [41] by Richards et al. concluded that opioid analgesics worsen attention but there was no statistically significant relationship between opioids and memory/executive function.
Three studies [30,41,44] that concluded that the use of opioids impairs cognitive function had limitations in study design, including but not limited to a small sample size, as described in Table 7. And none of them addressed driving function which was most used to assess cognition in other studies. The study by Cherrier et al. [30] has a limitation of study design, as the subjects who did not have chronic or significant daily pain were assessed shortly after a dose of opioid. In the study by Richards et al. [41], subjects used a relatively small dose of opioid. The study by Sjogren et al. [44] focused on subjects with mild pain intensity; potentially confounding factors such as anxiety or depression were not well addressed.

Studies focusing on ages over 65 years old
As it is described in Table 8, three [32,35,46] of four studies on the geriatric population concluded that using opioid analgesics does not necessarily worsen cognitive function. One study [30] concluded that opioids worsen cognitive function. Studies with large sample sizes [27,41] showed consistent results that using opioids does not worsen cognitive function.

Driving test results as a marker of cognitive/executive function
Five studies directly assessed driving function as described in Table 9.   The sample sizes of studies using driving function as an indicator for cognitive function were relatively small, and the information about pain type and intensity was limited. There was no consistency in the type of opioids studied. All articles assessing driving function concluded that there is no significant driving function decline attributable to opioid use.

Comparison of the results with other systematic reviews
Three other systematic reviews were consistent with our findings. A systematic review by Allegri et al. [48] investigated the neuropsychological effects of long-term use of opioids in patients with chronic noncancer pain. It concluded that opioids reduce attention, but there was no statistically significant difference noted in the other areas of cognitive function. Another systematic review by Ferreira et al. [49] assessed the impact of opioids on driving skills. They could not identify impaired simulated driving performance in subjects taking regularly scheduled opioids for symptom control. Pask et al. [50] concluded in their systematic review that in the 10 studies that met their inclusion criteria, six showed no consistent effect, and four indicated impairment of a range of cognitive function in patients receiving higher doses of opioids.

Conclusions Discussion
This study is one of few studies reviewing the impact on cognitive impairment of both pain and opioid pain medications. We found that opioid analgesics may impair cognitive function, but undertreated pain impairs cognitive function more. However, this systematic review has limitations.
Heterogenous studies were reviewed. Type of pain, opioids, and cognitive tests differ substantially between studies. This heterogeneity limited our ability to merge the findings of diverse reports. There were many cross-sectional studies and prospective observational cohort studies. Only one randomized controlled trial was identified in our review. Subgroup analyses were limited by the small number of the articles.
In the future, more studies are needed that focus on the over 65 years old population. Better quality study designs such as randomized controlled trials or well-designed cohort studies will provide a higher level of evidence. Further validation and uniform application of tools to test cognitive function will be required. Future studies should have suitably large sample sizes, and adjust for important confounding factors, such as mood disorder, level of education/intelligence, and underlying cognitive impairment, among others.

Conclusion and recommendation
In summary, evidence supports that chronic pain impairs cognitive function. In contrast, our identified studies found that using opioids does not significantly worsen cognition, including measures of executive function, memory, or driving ability.
Based on the present systematic reviews, we conclude that concerns over cognitive impairment by opioids should not prevent the treatment of moderate to severe pain and that untreated pain can impair cognitive function significantly. Careful use of opioids for moderate to severe pain does not necessarily worsen executive function, memory, or driving ability. However, our results cannot be generalized as they are derived from heterogeneous studies and cohorts. An open question remains whether nonsedating, nonopioid analgesics might become analgesics of choice to control pain without impairing cognition.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.