Metformin Use Is Not Associated With B12 Deficiency or Neuropathy in Patients With Type 2 Diabetes Mellitus in Qatar

Background Metformin may lead to B12 deficiency and neuropathy. There are no published data on the prevalence of Metformin-related B12 deficiency and neuropathy in the Arabian Gulf. Aims Determine whether Metformin intake is associated with B12 deficiency and whether B12 deficiency is associated with diabetic peripheral neuropathy (DPN) and painful diabetic neuropathy. Methods Patients with type 2 diabetes mellitus (T2DM) (n = 362) attending outpatient clinics at HMC underwent assessment of B12 levels, the DN4 questionnaire, and vibration perception threshold (VPT). Results Comparing Metformin to non-Metformin users there were no differences in B12 levels, VPT, or DN4. The prevalence of B12 deficiency (B12 <133 pmol/l) was lower (P < 0.01) in Metformin (8%) compared to non-Metformin (19%) users. Patients with B12 deficiency had a comparable prevalence and severity of sensory neuropathy and painful neuropathy to patients without B12 deficiency. Conclusion Serum B12 levels were comparable between Metformin and non-Metformin users with T2DM in Qatar. T2DM patients on Metformin had a lower prevalence of B12 deficiency. Furthermore, the prevalence and severity of neuropathy and painful diabetic neuropathy were comparable between patients with and without B12 deficiency.

Background: Metformin may lead to B12 deficiency and neuropathy. There are no published data on the prevalence of Metformin-related B12 deficiency and neuropathy in the Arabian Gulf.
aims: Determine whether Metformin intake is associated with B12 deficiency and whether B12 deficiency is associated with diabetic peripheral neuropathy (DPN) and painful diabetic neuropathy.
results: Comparing Metformin to non-Metformin users there were no differences in B12 levels, VPT, or DN4. The prevalence of B12 deficiency (B12 <133 pmol/l) was lower (P < 0.01) in Metformin (8%) compared to non-Metformin (19%) users. Patients with B12 deficiency had a comparable prevalence and severity of sensory neuropathy and painful neuropathy to patients without B12 deficiency.
conclusion: Serum B12 levels were comparable between Metformin and non-Metformin users with T2DM in Qatar. T2DM patients on Metformin had a lower prevalence of B12 deficiency. Furthermore, the prevalence and severity of neuropathy and painful diabetic neuropathy were comparable between patients with and without B12 deficiency.
Keywords: metformin, vitamin B12 deficiency, diabetic neuropathy, diabetic painful neuropathy, type 2 diabetes mellitus inTrODUcTiOn Metformin remains first-line therapy in type 2 diabetes mellitus (T2DM), with around 120 million users worldwide. It is increasingly used in overweight T2DM patients and those with polycystic ovary syndrome (1) Metformin therapy was shown to be associated with a significant reduction in the level of vitamin B12 over 50 years ago (2,3). A number of observational and placebo-controlled studies have confirmed that Metformin may reduce vitamin B12 levels (4)(5)(6)(7)(8)(9)(10). Indeed a recent study from Pakistan found that 29.7% of patients on Metformin had B12 deficiency (11) and another study from Brazil showed that B12 deficiency occurred in 22.4% of patients with T2DM on Metformin, and was further reduced in those on proton pump inhibitors (PPI)/ H2-antagonists (10). However, a recent meta-analysis showed that only 10/17 studies showed that Metformin use led to B12 deficiency and in four prospective studies B12 was reduced by approximately 57 pmol/L, within 6 weeks to 3 months of commencing Metformin (12).
A potential consequence of B12 deficiency is that it could directly result in neuropathy or exacerbate diabetic neuropathy. Indeed, the recent 2017 ADA position statement on diabetic neuropathy has emphasized the importance of excluding B12 deficiency in patients with diabetic neuropathy (13). However, there are conflicting reports on the association between Metformin-induced B12 deficiency and neuropathy, with some reports showing an association (14,15) while others have refuted this (11,(16)(17)(18). Furthermore, in a recent study from Turkey, while the prevalence of B12 deficiency was 38.4% there was no difference in B12 levels in those with and without neuropathy (19). Despite this there is wide spread administration of vitamin B12 therapy in patients in the Middle East and Far East, with a recent analysis from five teaching hospitals in Jordan, indicating that cyanocobalamin (B12), was the second most common injectable therapy after insulin (20). There are no published data on Metformin-related B12 deficiency or the relationship between B12 deficiency and diabetic neuropathy in the MENA region.
We have compared vitamin B12 levels in outpatients with T2DM in Qatar, in relation to Metformin use and further assessed for the prevalence and severity of painful neuropathy and sensory neuropathy in patients with B12 deficiency.

MaTerials anD MeThODs
Participants with T2DM (n = 362) were recruited from the National Diabetes & Endocrine Centers in Al-Wakra Hospital and Hamad General Hospital. The study was performed between March 6, 2017 and September 28, 2017.
Exclusion criteria included patients with a prior history of pernicious anemia, chronic kidney disease, previous bariatric surgery, gastrectomy, or small bowel resection for inflammatory bowel disease. This study was approved by the Institutional Review Board (IRB) of WCM-Q and HMC and all participants gave informed consent to take part in the study. The research adhered to the tenets of the declaration of Helsinki.

Demographic and Blood Measures
Data including age, duration of diabetes, blood pressure, body mass index (BMI), and medications including Metformin were recorded. HbA1c, lipid profile, renal function, and serum B12 were assessed.

B 12 assay
Blood was drawn directly into a dedicated evacuated tube (BD Diagnostic-Preanalytical Systems, Oxford, UK) and centrifuged at 3,500 g for 10 min and serum analyzed immediately or stored at −20°C until analysis on Beckman Dxi 600 (Beckman Coulter Inc., Brea, CA, USA). The Vitamin B12 assay is a competitive-binding immunoenzymatic assay. The amount of analyte in the sample was determined by means of a stored, multipoint calibration curve (Beckman Coulter Assay Manual 2015, Beckman Coulter Inc., Brea, CA, USA). Analytical sensitivity <50 pg/ml, traceability; traceable to an internal standard manufactured using the purified cyanocobalamin. Assay precision: 4.8-11.4%. B12 levels <133 pmol/l were considered deficient.

Diabetic Peripheral neuropathy (DPn) assessment
Vibration perception threshold (VPT) was measured on the pulp of the large toe with a Neurothesiometer (Horwell, Scientific Laboratory Supplies, Wilford, Nottingham, UK). The test was repeated three times and the average value was recorded. VPT at a cutoff point ≥15V was defined as DPN (21).

statistical analysis
Variables were compared between groups using a t-test and χ 2 test for continuous and categorical data, respectively. Data are expressed as mean (SD) of mean.
Univariate analysis by simple linear regression was applied to determine which variables are associated with B12 levels, VPT, and DN4 as outcome measures. Multiple linear regression analysis was used to determine the association between B12 levels, VPT, and DN4 after adjusting for confounding factors. Assumptions of linear regression were satisfied for normality, collinearity, and outliers. Additionally, residual plots were used to determine whether the models fit the assumptions.
association Between VPT, Dn4 score, and B 12 Simple linear regression analysis showed that VPT was positively associated with B12 (r = 0.18 P < 0.001). However, multiple linear regression analysis showed that this association was lost (β = 0.003, P = 0.25) after adjustment for confounding factors, including, age, diabetes duration, SBP, HbA1c, and PPI use. DN4 had no association with B12 levels.

DiscUssiOn
This is the first study from the Middle East region to assess the association between Metformin exposure and B12 levels and its relationship to diabetic neuropathy. We show no difference in B12 levels between Metformin and non-Metformin users and actually show that the prevalence of B12 deficiency was lower in patients on Metformin. This is in contrast to some but not all previously published studies (12). Furthermore, we show no difference in the prevalence of DPN or painful diabetic neuropathy in T2DM patients with and without B12 deficiency.
The 2018 ADA Clinical Practice Recommendations endorse screening Metformin users for vitamin B12 deficiency and the 2017 ADA diabetic neuropathy statement recommends that all patients with diabetic neuropathy should be assessed for B12 deficiency, to exclude a treatable cause of neuropathy (13). However, previous studies examining the relationship between Metformin use and B12 deficiency (12); and indeed between B12 deficiency and neuropathy have been conflicting (11,(14)(15)(16)(17)(18). Indeed, a study has shown a lower prevalence of DPN in T2DM patients on Metformin compared to those not on Metformin (6). Ahmed et al. (17) used the neuropathy total scoring system (NTSS) and showed that subjects with normal B12 levels had a comparable prevalence of DPN to those with low B12 levels (36.8 vs 32.3%), and no correlation between B12 levels and NTSS (17). Russo et al. compared 79 subjects with DPN and 184 without DPN and found no relationship to Metformin use (16). Chen et al. using a neurothesiometer and monofilaments in addition to a structured questionnaire also showed no relationship between Metformin use and peripheral neuropathy (25). In contrast Singh et al. showed that Metformin users had lower levels of B12 and a higher Toronto Neuropathy Scoring System (14). Roy et al. showed that patients on Metformin had a lower level of B12 and a reduction in median, ulnar, and peroneal nerve conduction (15). In the DPPOS study, while Metformin was associated with an increased risk of B12 deficiency, only 13 of the 56 participants on Metformin with low vitamin B12 had neuropathy, but there was no difference in neuropathy symptoms or the total Michigan Neuropathy Screening Instrument score (7). A recent study from India has shown an association between Metformin use and B12 levels as well as DPN assessed using the Toronto Clinical Scoring System and median, ulnar, peroneal, and posterior tibial nerve conduction velocity (26).
Given that we showed a lower prevalence of B12 deficiency in patients taking Metformin, we assessed confounding factors such as other medications, which may alter B12 levels. Sulfonylurea use in combination with Metformin is a significant independent risk factor for B12 deficiency (8). B12 levels have also been reported to be lower in older adults with prolonged PPI and H2 blocker use in one study (27) but not in another study (28). Vitamin B supplementation is prevalent in the Middle East and may also influence B12 levels (29,30). We show no association between B12 levels and concomitant use of sulfonylureas or calcium supplementations, but we do show small and significant associations with age, duration of diabetes, HbA1c, and treatment with vitamin D, PPI's, multivitamins, and B12. This is the first study to assess the relationship between Metformin use B12 deficiency and the prevalence and severity of DPN and diabetic painful neuropathy in Qatar. The prevalence of both sensory neuropathy and painful diabetic neuropathy was comparable to previously published data (31)(32)(33). A limitation of this study is that it is a retrospective cohort study, but VPT and DN4 were assessed without the investigators being aware of the treatment or B12 status.
The majority of patients had been prescribed Metformin as first-line therapy in accord with international guidelines, unless they were intolerant or it was withdrawn (34), and, therefore, it was not possible to recruit a larger number of patients not on Metformin. We cannot establish the exact duration of Metformin exposure, although we can assume that Metformin was prescribed shortly after diagnosis and, therefore, exposure is approximately equivalent to the duration of diabetes, which was approximately 10 years. As noted in the meta-analysis of Chapman et al. the B12 lowering effect of Metformin occurs within 6 weeks to 3 months of commencing Metformin (12).
In conclusion, we show no difference in B12 levels or the severity of DPN or painful diabetic neuropathy in Metformin compared to non-Metformin users. We also show no difference in vibration perception or painful diabetic neuropathy in those with and without B12 deficiency. These data urge the need for further larger, prospective studies to confirm or refute the current findings to support or challenge the highly prevalent practice of prescribing B12 for neuropathy across the Middle East.

eThics sTaTeMenT
This study was carried out in accordance with the recommendations of Investigator guidance: Investigator obligations (HRP-800), Institutional Review Board (IRB) of WCM-Q and HMC. The protocol was approved by the IRB of WCM-Q and HMC. All subjects gave written informed consent in accordance with the Declaration of Helsinki.
aUThOr cOnTriBUTiOns RM, TE, and GP had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: RM and TE. Acquisition, analysis, or interpretation of data: all authors. Drafting of the manuscript: RM, TE, and GP. Critical revision of the manuscript for important intellectual content: all authors. Statistical analysis: GP and RM. Administrative, technical, or material support: all authors.

acKnOWleDgMenTs
We thank all the participants for their efforts and commitment to be involved in the study. This work was supported by a grant from Qatar Foundation (BMRP 20038654).