Background/Objectives: It has been shown that the use of statins in patients with type 2 diabetes mellitus (T2DM) worsens hyperglycemia and hemoglobin A1c levels but may help in the preservation of pancreatic β-cell function. The potential role of a high pancreatic fat fraction (PFF) in this process has not yet been clarified. This study aimed to investigate whether the liver fat fraction (LFF) and PFF in T2DM patients is affected by statin therapy. Methods: This cross-sectional study involved a total of 140 T2DM patients, including both those who were receiving (n = 70) and those who were not receiving (n = 70) statin therapy. The mapping of the LFF and PFF utilizing the IDEAL-IQ sequence was conducted in magnetic resonance imaging. Results: In T2DM patients who used statins, the median PFF was higher compared to those who did not use statins (8.4 vs. 6.2%, p = 0.021), while the median LFF was found to be similar (8.4 vs. 8.9, p = 0.572). Variations in PFF were associated with the use of various statins (non-statin group: 6.2 vs. atovastatin: 8.7 vs. rosuvastatin: 3.2 vs. pitavastatin: 9.2, p = 0.004). The multivariable regression analysis indicated that insulin usage decreased log(LFF) by a factor of 0.16-fold (ꞵ ± SE = −0.16 ± 0.05, p = 0.010), and rosuvastatin usage reduced log(PFF) by 0.16-fold (ꞵ ± SE = −0.16 ± 0.07, p = 0.025), irrespective of other risk factors. Furthermore, the use of atorvastatin (ꞵ ± SE = 0.17 ± 0.06, p = 0.011) and pitavastatin (ꞵ ± SE = 0.19 ± 0.07, p = 0.008) were independently associated with an increase in log(PFF). Conclusions: In patients with T2DM, statin use did not show a significant effect on the liver fat fraction, but it caused differences in the pancreatic fat fraction. The observation of a lower pancreatic fat fraction in patients taking a rosuvastatin and atorvastatin dose of 40 mg/day suggests that different types and doses of statins may have varying effects on pancreatic fat accumulation.

Background and Objectives: Mild cognitive impairment (MCI) is heterogeneous within the Alzheimer’s disease (AD) continuum, and categorical labels may not reflect biological variability. We evaluated whether multimodal machine learning using routine clinical data and neuroimaging could support biologically informed enrichment across MCI and AD in a real-world memory clinic cohort. Methods: We analyzed 474 patients (1547 visits) with clinical and cognitive measures, laboratory parameters, MRI regional volumes, and FDG-PET regional uptake. Elastic Net and gradient boosting models were trained using nested cross-validation with strict patient-level separation. Results: Model discrimination improved as additional data modalities were added, and FDG-PET contributed the largest performance improvement. Hypometabolism in posterior default mode network regions consistently emerged as the most influential predictor. In the MCI subgroup, AD-like scores showed a continuous distribution consistent with biological enrichment. Conclusions: Multimodal models may provide an interpretable enrichment framework in heterogeneous memory clinic populations.