Trial of restarting and tolerating metformin (TreatMet).

This randomized, double-blind, placebo-controlled, n-of-1 crossover study assessed whether metformin's side effects are reproducible in patients with a history of metformin intolerance. Participants completed up to four cycles of 2 weeks of metformin exposure and 2 weeks of placebo exposure. Participants completed surveys based on the Gastrointestinal Symptom Rating Scale and the Treatment Satisfaction Questionnaire for Medication. The primary hypotheses were that treatment satisfaction would be equal for placebo and metformin and that more than 30% of the study enrollees would be able to adhere to a higher dose of metformin 6 months after participation. Thirteen patients (all women, mean age 52.4 years) enrolled, three of whom were lost to follow-up or were non-adherent to study protocol. Metformin was associated with significantly lower global treatment satisfaction scores compared with placebo (39.58 vs. 53.75, P < .05 ) but participants could not distinguish metformin from placebo and did not report higher rates of gastrointestinal side effects on metformin. Two out of 10 participants adhered to a higher dose of metformin after trial completion. Metformin appears to have barriers to use beyond its classic gastrointestinal side effects.

Carbohydrate-last meal pattern lowers postprandial glucose and insulin excursions in type 2 diabetes.

BACKGROUND: There are limited data regarding the timing of carbohydrate ingestion during a meal and postprandial glucose regulation. METHODS: Sixteen subjects with type 2 diabetes mellitus (T2DM) consumed the same meal on 3 days in random order: carbohydrate first, followed 10 min later by protein and vegetables; protein and vegetables first, followed 10 min later by carbohydrate; or all components together. Blood was sampled for glucose, insulin, glucagon-like peptide-1 (GLP-1), and glucagon measurements at baseline (just before meal ingestion) and subsequently at 30 min intervals up to 180 min. RESULTS: The incremental areas under the curve for glucose (iAUC0-180) and incremental glucose peaks were 53% and 54% lower, respectively, when carbohydrate was consumed last compared with carbohydrate consumed first (3124.7±501.2 vs 6703.5±904.6 mg/dL×180min, p<0.001; 34.7±4.1 vs 75.0±6.5 mg/dL, p<0.001) and 44% and 40% lower, respectively, compared with the all components together condition (3124.7±501.2 vs 5587.1±828.7 mg/dL×180min, p=0.003; 34.7±4.1 vs 58.2±5.9 mg/dL, p<0.001). Postprandial insulin excursions were lower (iAUC0-180: 7354.1±897.3 vs 9769.7±1002.1 µU/mL×min, p=0.003) and GLP-1 excursions higher (iAUC0-180: 3487.56±327.7 vs 2519.11±494.8 pg/mL×min, p=0.019) following the carbohydrate-last meal order compared with carbohydrate first. CONCLUSION: The carbohydrate-last meal pattern may be an effective behavioral strategy to improve postprandial glycemia.