Diabetes mellitus remains a major and expanding public-health problem. The International Diabetes Federation estimated that approximately 589 million adults were living with diabetes in 2024, with the burden concentrated heavily in low- and middle-income countries [1]. In India, where type 2 diabetes mellitus (T2DM) accounts for most diabetes care, long-term pharmacotherapy is therefore increasingly common and the adverse effects of chronic treatment have practical relevance beyond glycaemic control. Metformin is widely used because of its established glucose-lowering efficacy, tolerability and long clinical experience. A less visible consequence of prolonged exposure is a reduction in circulating vitamin B12. Randomised evidence has demonstrated a sustained fall in vitamin B12 concentrations during metformin treatment [2], while observational work has linked deficiency with both dose and duration of therapy [3]. Cross-sectional studies have also shown that vitamin B12 deficiency is not uncommon among patients with T2DM and may be more frequent in those receiving metformin [4]. Longer-term follow-up from the Diabetes Prevention Program Outcomes Study reinforced the association between cumulative metformin exposure and biochemical B12 deficiency [5]. The clinical importance is not restricted to an abnormal laboratory value. Vitamin B12 deficiency can produce haematological changes and neurological manifestations including paraesthesia, impaired vibration sense and peripheral neuropathy. These features overlap with diabetic neuropathy, so a reversible micronutrient deficiency may be missed when symptoms are attributed entirely to chronic hyperglycaemia. Current drug-safety guidance consequently advises testing vitamin B12 when deficiency is suspected and considering periodic monitoring in patients with risk factors, especially with higher doses or longer treatment duration [6]. The present study was undertaken to determine serum vitamin B12 status among T2DM patients receiving metformin and to compare it with that of patients managed without metformin. A secondary objective was to examine the relationship of vitamin B12 concentrations with diabetes duration and selected biochemical and treatment-related variables.

Study design and setting A cross-sectional comparative study was conducted from October 2021 to may 2022 at KAPV Government Medical College and hospital , Trichy . Study population The study included 90 participants with previously diagnosed T2DM. The metformin group comprised 45 patients who had been receiving metformin for at least 12 months. The comparison group comprised 45 patients with T2DM who had not received a biguanide, including metformin, during the preceding year and were managed with insulin and/or other oral hypoglycaemic agents. Eligibility criteria Adults of either sex with previously diagnosed T2DM were eligible. Exclusion criteria were a history of anaemia, previous blood transfusion, renal insufficiency, thyroid disease, alcohol intake, previous gastric surgery, proton-pump inhibitor use, or a malabsorption syndrome. Sample collection and biochemical analysis A 3 mL venous blood sample was collected from the antecubital vein in a plain tube. Serum was separated by centrifugation at 3000 rpm for 15 minutes within 2 hours of collection and stored at -20°C until analysis. Random plasma glucose was estimated by the glucose oxidase-peroxidase end-point method (Erba). Serum urea was measured by the urease-GLDH kinetic UV method (Accucare), and serum creatinine by the Jaffe initial-rate method (Erba). Serum vitamin B12 was measured using a sandwich ELISA. The assay used vitamin B12-specific antibody-coated microplate wells, a biotin-labelled detection system and horseradish peroxidase-avidin, with tetramethylbenzidine as substrate. Absorbance was measured at 450 nm. The calibration range described for the assay was 15.6-1000 pmol/L, with sample dilution performed when concentrations were outside the calibration range. For categorical reporting in this study, vitamin B12 <240 pmol/L was considered deficient, 240-900 pmol/L was considered within the stated reference interval, and values >900 pmol/L were categorised as high. Statistical analysis Continuous variables were summarised as mean and standard deviation. Between-group comparisons were performed using the independent-samples Student t test. Pearson correlation coefficients were calculated between serum vitamin B12 and clinical or biochemical variables within each study group. A p value <0.05 was considered statistically significant. The categorical vitamin B12 distribution is reported descriptively because no inferential test for that distribution was documented. Ethical considerations The study was approved by the Institutional Ethical Committee of Government Kilpauk Medical College, Chennai. The study was explained to participants and written informed consent was obtained before enrolment and blood collection.

Ninety participants were analysed, 45 in the metformin group and 45 in the non-metformin comparison group. The metformin group was older than the comparison group (49.96 ± 12.019 versus 43.31 ± 10.377 years; p=0.006). Duration of diabetes was similar between groups (5.5111 ± 3.08679 versus 5.1333 ± 4.27253 years; p=0.632). Random plasma glucose and serum urea also did not differ significantly. Serum creatinine was lower in the metformin group (0.8663 ± 0.40739 mg/dL) than in the non-metformin group (1.0211 ± 0.26128 mg/dL; p=0.035) (Table 1 and Figure 3). Serum vitam in B12 showed the clearest between-group difference. Mean vitamin B12 was 754.504 ± 395.7267 pmol/L in the metformin group compared with 1111.202 ± 414.7199 pmol/L in the non-metformin group (p<0.001) (Table 1 and Figure 2). When categorised using the study thresholds, 7 metformin-treated participants were vitamin B12 deficient, compared with 3 participants in the non-metformin group. The corresponding distributions across deficient, normal and high categories are shown in Table 2 and Figure 1. Within the non-metformin group, vitamin B12 was inversely correlated with duration of diabetes (r=-0.449, p=0.002); correlations with glucose, urea, creatinine, age, insulin dose, insulin duration and other drugs were not statistically significant (Table 3). In the metformin group, vitamin B12 was inversely correlated with duration of diabetes (r=-0.339, p=0.023). The variable labelled “Metformin” in the original correlation table also showed an inverse correlation with vitamin B12 (r=-0.572, p<0.001), but its operational definition was not documented; it therefore cannot be interpreted specifically as dose, treatment duration or cumulative exposure (Table 4 and Figure 4). Other correlations in this group were not statistically significant. Table 1. Clinical and biochemical comparison between study groups (n=90). Variable Non-metformin group (n=45) Metformin group (n=45) p value Age (years) 43.31 ± 10.377 49.96 ± 12.019 0.006 Duration of diabetes (years) 5.1333 ± 4.27253 5.5111 ± 3.08679 0.632 Random plasma glucose (mg/dL) 138.242 ± 25.9523 142.331 ± 41.2443 0.575 Serum urea (mg/dL) 27.160 ± 6.3434 28.178 ± 8.7029 0.528 Serum creatinine (mg/dL) 1.0211 ± 0.26128 0.8663 ± 0.40739 0.035 Serum vitamin B12 (pmol/L) 1111.202 ± 414.7199 754.504 ± 395.7267 <0.001 Values are mean ± standard deviation. p values are from independent-samples t tests. Table 2. Distribution of serum vitamin B12 categories by treatment group. Vitamin B12 category Non-metformin group Metformin group Total Deficient (<240 pmol/L) 3 7 10 Normal (240-900 pmol/L) 9 26 35 High (>900 pmol/L) 33 12 45 Total 45 45 90 Counts are reported directly; no categorical inferential test was documented. Table 3. Correlation of serum vitamin B12 with selected variables in the non-metformin group. Variable Pearson r p value n Random plasma glucose 0.067 0.660 45 Serum urea -0.010 0.945 45 Serum creatinine 0.184 0.225 45 Age -0.281 0.061 45 Duration of diabetes -0.449 0.002 45 Insulin dose -0.097 0.585 34 Insulin duration 0.166 0.357 33 Other drug variable 0.310 0.303 13 Pearson correlation; p<0.05 considered statistically significant. Table 4. Correlation of serum vitamin B12 with selected variables in the metformin group. Variable Pearson r p value n Random plasma glucose -0.176 0.248 45 Serum urea 0.167 0.273 45 Serum creatinine 0.021 0.890 45 Age -0.287 0.056 45 Duration of diabetes -0.339 0.023 45 Metformin (as labelled) -0.572 <0.001 45 The original correlation table labels the final variable simply as “Metformin”. Its operational definition was not specified; accordingly, the coefficient should not be interpreted specifically as metformin dose, treatment duration or cumulative exposure. Negative r values indicate inverse associations. The “Metformin” variable is reproduced as labelled in the original analysis; its operational definition was not specified.

The principal finding of this study was a marked difference in serum vitamin B12 between treatment groups. Participants receiving metformin had a mean concentration of 754.504 ± 395.7267 pmol/L, whereas those managed without metformin had a mean concentration of 1111.202 ± 414.7199 pmol/L. The difference was highly significant (p<0.001). The categorical data pointed in the same direction: 7 of 45 metformin-treated participants met the study definition of deficiency, compared with 3 of 45 participants in the non-metformin group. Because no categorical hypothesis test was reported, that difference is best interpreted descriptively rather than as proof of a statistically significant difference in deficiency prevalence. This pattern is consistent with population-based evidence. Reinstatler and colleagues found biochemical B12 deficiency more often among metformin users than non-users in the US National Health and Nutrition Examination Survey [7]. Nervo et al. similarly documented a high frequency of low vitamin B12 among metformin-treated patients in Brazil [8], while a case-control study from Pakistan reported an appreciable burden of deficiency among metformin-treated T2DM patients [9]. These studies vary in their diagnostic cut-offs, background diet and patient characteristics, which partly explains why prevalence estimates are not uniform across settings. Indian data are particularly relevant because nutritional intake, diabetes duration and access to laboratory monitoring can differ substantially across clinical populations. Singh et al. reported significantly lower vitamin B12 concentrations and greater clinical neuropathy scores in metformin-exposed patients, with a strong inverse relationship between cumulative metformin exposure and B12 [10]. The present study did not formally grade neuropathy, but its biochemical findings reinforce the practical concern that B12 deficiency may coexist with, and potentially be mistaken for, diabetic peripheral neuropathy. The variable labelled “Metformin” in the original analysis was inversely correlated with serum B12 (r=-0.572, p<0.001). Because the underlying operational definition of this variable was not documented, the present coefficient cannot be treated as evidence of a dose-response, treatment-duration or cumulative-exposure relationship. Nevertheless, the direction of the association is compatible with earlier work in which Ting et al. identified both current metformin dose and treatment duration as risk factors for vitamin B12 deficiency [3], while Beulens et al. reported lower cobalamin status with longer or more intensive metformin use [11]. Long-term trial data also indicate that B12 concentrations may decline during continued metformin therapy [2,5]. Separately, duration of diabetes was inversely correlated with B12 in both study groups, indicating that factors other than metformin may also contribute to lower B12 status. Other Indian studies have reported broadly similar findings. Roy et al. described vitamin B12 deficiency and peripheral neuropathy in metformin-treated patients with early T2DM [12]. Raizada et al. found lower serum B12 concentrations in patients receiving metformin than in those never exposed to the drug after accounting for diabetes duration [13]. Gupta et al. also observed an inverse association between duration of metformin use and B12 concentrations, together with neuropathy findings [14]. More recently, an Indian analysis using a metformin usage index found deficiency to be related to cumulative metformin exposure [15]. Collectively, these studies support a biological gradient rather than a simple binary drug effect. The clinical implication is straightforward but should be framed cautiously. Serum B12 should not be interpreted in isolation from symptoms, diet, haematological indices and other causes of malabsorption. Nonetheless, an unexplained sensory neuropathy, macrocytosis or anaemia in a person receiving long-term metformin warrants consideration of B12 deficiency. Functional markers can add information when serum B12 is equivocal. In the HOME trial post hoc analysis, metformin increased methylmalonic acid over time and the rise was associated with worsening neuropathy scores, supporting the clinical relevance of tissue-level deficiency [16]. Current regulatory guidance similarly recommends testing when deficiency is suspected and considering periodic monitoring in higher-risk metformin users [6]. The age difference between study groups deserves attention. The metformin group was older, despite the original protocol describing the groups as comparable, and age showed a negative but non-significant correlation with B12 within each group. Serum creatinine also differed between groups. These imbalances, together with the absence of multivariable adjustment, mean that the observed between-group B12 difference cannot be attributed to metformin alone. The cross-sectional design establishes association rather than temporality or causality. There were further limitations. The study was single-centre and included only 90 participants. Vitamin B12 status was assessed by serum concentration without holotranscobalamin, methylmalonic acid or homocysteine. Dietary B12 intake, vegetarian status, haematological indices and a standardized neuropathy score were not reported in the analysed results. The exact operational definition of the recorded metformin-exposure variable used in the correlation table was also not specified. Finally, the use of a locally stated B12 cut-off limits direct comparison with studies that define deficiency at lower thresholds. These limitations should temper generalisation, but they do not negate the consistent finding of lower serum B12 among metformin-treated participants.

In this cross-sectional comparative study, T2DM patients receiving metformin had significantly lower mean serum vitamin B12 concentrations than patients managed without metformin, and a greater number met the study definition of vitamin B12 deficiency. Vitamin B12 also showed an inverse relationship with diabetes duration in both groups. An additional inverse correlation was observed with the variable labelled “Metformin” in the metformin group, but its operational definition was not documented. The findings support an association between sustained metformin therapy and lower vitamin B12 status, while the study design, baseline group differences and absence of adjusted analysis preclude a causal conclusion. Periodic clinical assessment for B12 deficiency is reasonable in long-term metformin users, particularly when neuropathic or haematological features develop

1. International Diabetes Federation. IDF Diabetes Atlas. 11th ed. Brussels, Belgium: International Diabetes Federation; 2025. 2. de Jager J, Kooy A, Lehert P, Wulffelé MG, van der Kolk J, Bets D, et al. Long term treatment with metformin in patients with type 2 diabetes and risk of vitamin B-12 deficiency: randomised placebo controlled trial. BMJ. 2010;340:c2181. doi:10.1136/bmj.c2181. 3. Ting RZ, Szeto CC, Chan MH, Ma KK, Chow KM. Risk factors of vitamin B12 deficiency in patients receiving metformin. Arch Intern Med. 2006;166(18):1975-1979. doi:10.1001/archinte.166.18.1975. 4. Pflipsen MC, Oh RC, Saguil A, Seehusen DA, Topolski R. The prevalence of vitamin B12 deficiency in patients with type 2 diabetes: a cross-sectional study. J Am Board Fam Med. 2009;22(5):528-534. doi:10.3122/jabfm.2009.05.090044. 5. Aroda VR, Edelstein SL, Goldberg RB, Knowler WC, Marcovina SM, Orchard TJ, et al. Long-term metformin use and vitamin B12 deficiency in the Diabetes Prevention Program Outcomes Study. J Clin Endocrinol Metab. 2016;101(4):1754-1761. doi:10.1210/jc.2015-3754. 6. Medicines and Healthcare products Regulatory Agency. Metformin and reduced vitamin B12 levels: new advice for monitoring patients at risk. Drug Safety Update. 2022;15(11):1. 7. Reinstatler L, Qi YP, Williamson RS, Garn JV, Oakley GP Jr. Association of biochemical B12 deficiency with metformin therapy and vitamin B12 supplements: the National Health and Nutrition Examination Survey, 1999-2006. Diabetes Care. 2012;35(2):327-333. doi:10.2337/dc11-1582. 8. Nervo M, Lubini A, Raimundo FV, Faulhaber GA, Leite C, Fischer LM, et al. Vitamin B12 in metformin-treated diabetic patients: a cross-sectional study in Brazil. Rev Assoc Med Bras (1992). 2011;57(1):46-49. doi:10.1590/S0104-42302011000100015. 9. Iftikhar R, Kamran SM, Qadir A, Iqbal Z, bin Usman H. Prevalence of vitamin B12 deficiency in patients of type 2 diabetes mellitus on metformin: a case control study from Pakistan. Pan Afr Med J. 2013;16:67. doi:10.11604/pamj.2013.16.67.2800. 10. Singh AK, Kumar A, Karmakar D, Jha RK. Association of B12 deficiency and clinical neuropathy with metformin use in type 2 diabetes patients. J Postgrad Med. 2013;59(4):253-257. doi:10.4103/0022-3859.123143. 11. Beulens JWJ, Hart HE, Kuijs R, Kooijman-Buiting AMJ, Rutten GEHM. Influence of duration and dose of metformin on cobalamin deficiency in type 2 diabetes patients using metformin. Acta Diabetol. 2015;52(1):47-53. doi:10.1007/s00592-014-0597-8. 12. Roy RP, Ghosh K, Ghosh M, Acharyya A, Bhattacharya A, Pal M, et al. Study of vitamin B12 deficiency and peripheral neuropathy in metformin-treated early type 2 diabetes mellitus. Indian J Endocrinol Metab. 2016;20(5):631-637. doi:10.4103/2230-8210.190542. 13. Raizada N, Jyotsna VP, Sreenivas V, Tandon N. Serum vitamin B12 levels in type 2 diabetes patients on metformin compared to those never on metformin: a cross-sectional study. Indian J Endocrinol Metab. 2017;21(3):424-428. doi:10.4103/ijem.IJEM_529_16. 14. Gupta K, Jain A, Rohatgi A. An observational study of vitamin B12 levels and peripheral neuropathy profile in patients of diabetes mellitus on metformin therapy. Diabetes Metab Syndr. 2018;12(1):51-58. doi:10.1016/j.dsx.2017.08.014. 15. Shivaprasad C, Gautham K, Ramdas B, Gopaldatta KS, Nishchitha K. Metformin usage index and assessment of vitamin B12 deficiency among metformin and non-metformin users with type 2 diabetes mellitus. Acta Diabetol. 2020;57(9):1073-1080. doi:10.1007/s00592-020-01526-4. 16. Out M, Kooy A, Lehert P, Schalkwijk CA, Stehouwer CDA. Long-term treatment with metformin in type 2 diabetes and methylmalonic acid: post hoc analysis of a randomized controlled 4.3-year trial. J Diabetes Complications. 2018;32(2):171-178. doi:10.1016/j.jdiacomp.2017.11.001.