Regular use of opioids and dementia, cognitive measures, and neuroimaging outcomes among UK Biobank participants with chronic non-cancer pain

Abstract

Introduction: We investigated the association between regular opioid use and incident dementia, neuroimaging outcomes, and cognitive measures.

Methods: Cox regression was used to assess the association between opioid use and incident dementia among197,673 UK Biobank participants with chronic non-cancer pain. Linear and logistic regression were applied to explore the associations between opioid use and dementia-related neuroimaging and cognitive function outcomes.

Results: Regular opioid use was associated with a 20% higher risk of all-cause dementia and a 49% higher risk of vascular dementia (VD) compared with those not using analgesics. Moreover, those using strong opioids had a 72% higher risk of all-cause dementia and a 155% higher risk of VD. Strong opioid use was also linked to reductions in hippocampal, white matter, and total brain volumes. Lastly, regular opioid use was associated with lower fluid intelligence.

Discussion: A higher risk of dementia was observed among participants regularly using opioids, escalating with opioid strength.

Highlights: Regular opioid use was associated with an increased risk of all-cause dementia and VD. Those using strong opioids had a much higher risk of all-cause dementia and VD. Strong opioid use was also associated with worse neuroimaging outcomes. Regular opioid use was also associated with lower fluid intelligence.

Keywords: UK Biobank; cognitive function; dementia; neuroimaging; opioid.

FIGURE 1

Flow chart of study participants.

FIGURE 2

Forest plot of the association between opioid use and neuroimaging outcomes (z‐score). ^* All neuroimaging outcomes for each individual were normalized by head size and converted to z‐scores based on the mean and standard deviation. ^† Non‐opioid analgesic use included NSAIDs or acetaminophen use.^‡ The analgesic exposure was categorized into three groups (non‐analgesic use, non‐opioid analgesic use and opioid use) in the primary analysis models and four groups (non‐analgesic use, non‐opioid analgesic use, weak opioid use, and strong opioid use) in the secondary analysis models, respectively. ^§ Model was adjusted for age, sex, education, ethnicity, residential place and socioeconomic status, living alone status, smoking status, alcohol drinking status, physical activities, BMI categories, sleeping duration categories, family visits and social activities engagement, major surgery history, multimorbidity conditions, total number of sites of chronic pain (head, face, neck/shoulder, back, abdomen, hip, and knee) or general widespread pain, acetylsalicylic acid use, glucocorticoid use, anticholinergic use and APOE ε4 carrier status. ^|| The value is approximately equal to 0 but lower than 0. APOE, apolipoprotein E; BMI, body mass index; NSAIDs, non‐steroidal anti‐inflammatory drugs.