The purpose of this study is to determine whether exposure to prescription opioids is associated with brain structural changes on magnetic resonance imaging (MRI), with the primary outcome of white matter integrity as measured by fractional anisotropy of the corpus callosum.

The investigators will recruit 200 adults aged 65 years and older with chronic pain of the knee or shoulder. Each will receive an individualized multimodal pain treatment pathway for 6-weeks - for half of participants this will include a prescription opioid; the other half will receive an identical placebo. After 6 weeks, participants will undergo opioid/placebo taper. The investigators will obtain neuroimaging (MRI), psychometric/cognitive testing, pain scores, and several exploratory patient-reported outcomes at several time-points: before the intervention (all tests), at 6 weeks (all tests except MRI), 6-month phone call (only phone based patient reported outcomes- no MRI), and at 1 year (all tests).

Inclusion Criteria:

• 65 years of age or older
• Pain for longer than 3 months (chronic)
• Moderate to severe pain nearly every day (averaged PEG score >5)
• Persistent pain despite use of >1 analgesic (Tylenol, Ibuprofen, lidocaine patch, etc)

Exclusion Criteria:

• Untreated severe/uncontrolled Mental health (schizophrenia, bipolar disorder, psychosis, history of suicidality, depression or anxiety, PTSD)
• Severe liver (cirrhosis) or kidney disease (ESRD on dialysis or GFR< 30)
• Cognitive impairment (Kokmen Short Test of Mental Status score < 29)
• Life expectancy < 12 months
• Contraindications to MRI
• Non-English speaking
• Opioid use within the past 6 months
• Contraindications or previous intolerance to prescription opioids
• Substance use disorder or high-risk for opioid-related adverse effects (Opioid Risk Tool ≥ 8)
• Residence outside of Minnesota or Wisconsin during opioid intervention period
• Another pain generator that is greater than their knee or shoulder or low back pain
• Anticipated surgery for joint replacement (Knee or shoulder) sooner than 3 months.

Deidentified summary measures will be made openly available to facilitate future scientific research from this NIA-supported effort. Deidentified individual level data will be available through a controlled access plan, which will require a confidential agreement and institutional review board (IRB) approvals as described in Element 5. MRI data will be linked to the cognitive measures as well as demographic information and clinical measures. In addition to numeric MRI outputs, raw or preprocessed MRI data will be made available in accordance with Brain Imaging Data Structure (BIDS) standards. Non-imaging data including psychometric testing and patient questionnaires will be available in raw form.

Scientific data will be deposited into the repository at the time of publication or the end of the award period, which comes first. Data will be preserved for a minimum of 5 years following the end of the grant period.

De-identified datasets from this research study will be made available in accordance with the NIH Aging Research Biobank. Additional requests will be considered to the extent permissible under our IRB approvals, and local, state and federal laws and regulations, including the Privacy Rule. Additional requests for data should be directed to the study's Principal Investigators, who will evaluate the requests in a timely manner. Incoming proposals will be screened to ensure that the requests did not overlap or conflict with the funded efforts of this project. Approved requests will require a signed data use agreement and IRB approval to ensure that the usual procedures for participant protection are followed.

All parties are masked other than the individual that will be dispensing the drug and preparing the medication (Placebo vs. Opioid).