Background/Objectives: Metabolic dysfunction-associated steatotic liver disease (MASLD) has been linked to dietary fat quality and polyunsaturated fatty-acid metabolism. We evaluated whether dietary n-6 fatty-acid intake, serum dihomo-γ-linolenic acid (DGLA), and desaturase-based indices for Δ5-desaturase (D5D) and Δ6-desaturase (D6D) are associated with MASLD. Methods: We conducted a cross-sectional analysis within the HOZUGAWA health checkup cohort in Japan (n = 289; 100 MASLD, 189 non-MASLD). Participants underwent hepatic ultrasonography, dietary assessment using the Brief Self-Administered Diet History Questionnaire, and fasting serum metabolomics by gas chromatography–mass spectrometry with solid-phase dehydration derivatization. Enzyme indices were defined as the D5D index = arachidonic acid/DGLA and the D6D proxy index = DGLA/linoleic acid (hereafter referred to as the D6D index) because γ-linolenic acid was not measured. Natural-log-transformed D5D index, D6D index, DGLA, and total dietary n-6 fatty-acid intake were entered into multivariable logistic regression models for MASLD adjusted for age, sex, BMI, alcohol intake, and total energy. Results: Compared with non-MASLD, MASLD showed higher serum DGLA, lower D5D index, and higher D6D index (all p ≤ 0.005), with no between-group differences in total energy intake, linoleic acid, total polyunsaturated fatty acids, or total dietary n-6 fatty-acid intake. Higher ln D5D was independently associated with lower odds of MASLD (OR 0.62, 95% CI 0.42–0.86), whereas higher ln D6D index (OR 1.42, 95% CI 1.04–1.95) and ln DGLA (OR 1.62, 95% CI 1.13–2.43) were each positively associated. Total dietary n-6 fatty-acid intake was not independently associated with MASLD. Conclusions: In this Japanese health examination cohort, an imbalance in estimated desaturase activities—lower D5D index and higher D6D index—together with higher serum DGLA was independently associated with MASLD, whereas n-6 intake showed no group difference or independent association. These findings suggest that enzyme-linked endogenous n-6 metabolic status may be more closely related to the MASLD phenotype than intake quantity alone.

Background/Objectives: Postprandial glucose variability is a key challenge in diabetes management for patients receiving multiple daily insulin injections (MDI). This study evaluated transformer-based machine-learning models for predicting post-prandial glucose peaks and nadirs using pre-meal glucose, insulin dose, and nutritional input. Methods: In this observational study, 58 adults with diabetes provided dietary records, insulin logs, and continuous glucose monitoring data. After preprocessing and participant-level splitting (64:16:20), model-ready datasets comprised 6155/1449/1805 (train/validation/test) meal events for the Full-Nutrition model and 6299/1484/1849 for the Carbohydrate and Available-Carbohydrate models. We evaluated three transformer-based models and assessed performance using MAE, R², and the Clarke error grid. Results: The Full Nutrition Model achieved MAEs of 32.2 mg/dL (peak) and 21.8 mg/dL (nadir) with R² values of 0.58 for both. Carbohydrate-based models showed similar accuracy. Most predictions fell within Clarke error grid Zones A and B. Conclusions: Transformer-based machine-learning models can accurately predict postprandial glucose variability in MDI-treated patients. Carbohydrate-only inputs performed comparably to full-nutrient data, supporting the feasibility of simplified dietary inputs in clinical applications.