UC Irvine

Background. Metformin is the most common first-line medication for type 2 diabetes, and long-term use has been repeatedly linked to decreased vitamin B12 absorption, but the strength of this link has not been fully re-evaluated in more recent nationally representative U.S. data.Methods. This cross-sectional study analyzed pooled data from the National Health and Nutrition Examination Survey (NHANES) 2011–2014 among U.S. adults with diagnosed diabetes (N = 1,426). Vitamin B12 deficiency was defined as serum B12 <200 pg/mL among the 1,278 participants with a valid B12 measurement. Metformin use was identified from prescription medication data and duration of use was evaluated among current users. Logistic regression models were adjusted for demographic, clinical, and lifestyle covariates and accounted for the complex NHANES survey design; missing covariate data was addressed with multiple imputation. Sensitivity analyses used methylmalonic acid (MMA) alone as an alternative deficiency marker and excluded participants likely to have type 1 diabetes. Results. After adjustment, current metformin users had significantly higher odds of biochemical B12 deficiency than non-users (adjusted OR = 2.49, 95% CI: 1.22–5.07, p = 0.012). The odds of deficiency among users increased by 9% for every additional year of metformin use (adjusted OR = 1.09, 95% CI: 1.02–1.18), indicating a dose-dependent effect. This association was not observed when deficiency was defined using MMA alone (OR = 0.89, 95% CI: 0.69–1.16). Restricting the sample to exclude likely type 1 diabetes cases did not significantly change the primary result (adjusted OR = 2.46, 95% CI: 1.20–5.03). Conclusion. These results are consistent with the biochemical association between metformin use and vitamin B12 deficiency seen in previous NHANES-based research, extended to a more recent survey period, and confirm that duration of use should be considered clinically relevant. Periodic B12 screening in adults on long-term metformin therapy is considered to be a reasonable, low-cost practice that current evidence suggests is not yet reliably implemented in routine care.