This Phase II pragmatic hybrid effectiveness-implementation trial tests whether acarbose, titrated using continuous glucose monitoring (CGM) to blunt post-prandial excursions, reduces 4-week pain area-under-the-curve (AUC) versus placebo in adults with painful diabetic peripheral neuropathy (DPN) and high glycemic variability. Secondary objectives assess CGM variability metrics, microvascular reactivity, inflammatory markers, safety, and feasibility of a pharmacist-led titration workflow using loaner CGMs across multi-region community clinics.

Adults with T2D, painful DPN, and high CGM variability (e.g., MAGE >50 mg/dL on run-in) are randomized 1:1 to acarbose vs matching placebo for 4 weeks, on stable background analgesics. A standardized pharmacist-led algorithm escalates acarbose to target post-prandial spikes, guided by blinded CGM trend review. The primary endpoint is 4-week daily pain AUC captured via ePRO. Key secondary endpoints include changes in MAGE and time-in-range (TIR), skin microvascular reactivity (laser speckle), serum IL-6, patient global impression of change, rescue-analgesic use, and adverse events. Implementation outcomes (acceptability, feasibility, adoption, cost) are collected to inform scale-up in HIC and LMIC community settings.

Inclusion:

• Age 18-75 years.
• Type 2 diabetes ≥1 year; HbA1c 7.0-10.0% within 8 weeks of randomization.
• Painful DPN meeting clinical criteria; average daily pain NRS ≥4 during run-in.
• High CGM variability on 7-10 day run-in (e.g., MAGE >50 mg/dL).
• Stable analgesic regimen ≥4 weeks pre-baseline.
• Able to use CGM and ePRO; provides informed consent.

Exclusion:

• Type 1 diabetes; non-diabetic neuropathies.
• Contraindications to acarbose (e.g., chronic intestinal malabsorption, inflammatory bowel disease).
• eGFR <45 mL/min/1.73 m²; significant hepatic disease (ALT/AST >3× ULN).
• Use of α-glucosidase inhibitors within 3 months.
• Recent change (<3 months) in GLP-1/GIP agonists, SGLT2i, or basal/bolus insulin strategy.
• Pregnancy/lactation; other conditions compromising safety/assessments.

De-identified individual participant data underlying the results reported in this study, including the analyzable dataset and data dictionary, will be made available.

Data will be available beginning 6 months after publication and for up to 5 years.

Data will be made available to qualified researchers upon reasonable request following publication of the primary results. Requests must include a scientifically sound proposal and statistical analysis plan. Access will require approval by the sponsor and execution of a data use agreement.