Background  As one of most widely-used biguanides, metformin can induce the lactic acidosis in patients with renal failure though its incidence is very low. However, lactic acidemia induced by metformin was reported in patients without renal dysfunction. It is unclear that whether lactatemia exists in diabetic patients with normal renal function in Chinese or not and its influencing factors. This study aimed to clarify the influencing factors of lactic acid, and identify a practiced clinical marker to predict the hyperlactacidemia in diabetics with normal renal function.
Methods  The clinical data and venous blood samples of 1024 type 2 diabetic patients treated with (n=426) or without metformin (n=599) were collected. The lactic acid was assayed by enzyme-electrode method. The biochemical indexes included creatinine (Cr) and hepatase were measured with enzymatic procedures. The lactic acid concentrations of different Cr subgroups were compared, and the correlation and receiver operating characteristic curve analysis were used.
Results  The mean lactic acid level and the proportion of hyperlactatemia of metformin group were significantly higher than that of non-metformin group (P <0.01), but no lactic acidosis was found in all patients. The correlation and multiple stepwise regression analysis indicated that the correlative factors of lactic acid in turn were Cr, metformin, alanine transferase (ALT), body mass index (BMI), Urine albumin (Ualb), and blood urea nitrogen (BUN) in total patients; and Cr, ALT, BMI and BUN in non-metformin treated patients; Cr and ALT in metformin-group. The lactate concentration increased with the increment of Cr levels, and reached its peak at Cr 111–130 µmol/L, and the optimal cutoff of Cr in predicting hyperlactacidemia was 96.5 µmol/L.
Conclusions  Metformin can increase the incidence of lactatemia in type 2 diabetic patients without renal dysfunction. Cr, ALT, and BMI are independent associated factors of blood lactic acid levels. There is low proportion of lactatemia in type 2 diabetics without metformin therapy, the optimal cutoff of Cr to predict lactatemia in these patients is 96.5 µmol/L.

·LogIn/LogOut

·Fulltext PDF(189K) Free

·Abstract download TXT | XML

·Articles in CMJ by LIU Fang LU Jun-xi

·Articles in PubMed by LIU F LU JX

·Put into my bookshelf

·Email to Friend

·Email to author

·Visit:1271

·Download:608

·Advanced Search

·Related Articles

·Change font size:

·Cannot read some characters

Serum lactate level is a balance between its synthesis and elimination. The lactic acid was produced from muscle, adipose tissue, and liver through anoxic metabolism of glucose, and mainly breakdown in liver through gluconeogenesis. It was excreted almost completely from urine by kidney. Metformin, one of most widely-used biguanides, was used as first-line choice for type 2 diabetic patients in the treatment guideline of 2006 International Diabetes Federation (IDF), American diabetes Association (ADA), and 2007 edition Guideline of Chinese Diabetes Society (CDS).¹ One serious adverse effect of biguanides was lactic acidosis, but the incidence of lactic acidosis induced by metformin was rare and lower than that by phenformin,² a drug also from the biguanide class. Reports on lactic acidosis induced by metformin happened mainly in patients whose renal function was impaired, or patients who had alcoholic addiction, or gastrointestinal disease, or with oxygen deficiency diseases such as congestive heart failure, chronic obstructive pneumonic disease.^2-4 One of the contraindications of metformin is renal dysfunction with either glomerular filtration ratio (GFR) <60 ml/min or plasma creatinine (Cr) over 107 µmol/L or 130 µmol/L varying in different countries.^1,3-5 Recently, studies from Abbasi et al⁶ and Khan et al⁷ reported that metformin could caused lactatemia in type 2 diabetic patients with normal renal function. However, others considered that metformin therapy was not associated with an increased risk of lactic acidosis compared with other anti-hyperglycemic treatment if the agent was prescribed under indications.^3-6 Therefore, it is necessary to identify whether lactatemia exist in diabetic patients with normal renal function and the influence of metformin on the lactic acid concentrations in these populations. It was well known that the acknowledged indicator of renal function was GFR, but its determination was complex and expensive. Thus it is necessary to choose a practiced clinical index to predict the risk of lactatemia. The object of the present study was: (1) to find the influencing factors of lactic acid concentrations in type 2 diabetic patients treated with or without metformin; (2) to identify a practiced clinical index to predict the risk of lactatemia.

Subjects and grouping
Totally 1024 patients aged from 25 to 86 years were recruited from the in-patients of Shanghai Clinical Center of Diabetes from October 2005 to October 2008. The inclusion criteria were as follows: (1) diagnosed as type 2 diabetes by 1999 WHO criteria; (2) the renal function was normal or only slightly impaired, indicated by GFR >60 ml/min and Cr below 130 µmol/L; (3) do physical activity freely. Participants with the following conditions were excluded: (1) chronic liver disease, (2) congestive heart failure (CHF), (3) severe chronic obstructive pneumonic disease (COPD), (4) alcohol addiction, and (5) chronic diarrhea and innutrition state. The study was approved by the Ethics Committee of Shanghai 6th People′s Hospital. Written informed consents were obtained from all subjects. We enrolled 425 patients who were treated with metformin 500 mg three times daily for at least 1 week, alone or combined with other glucose lowering agents (metformin-treated group), and 599 patients whose glucose level were controlled by sulphanylureas, insulin or other anti-hyperglycemic drugs but not metformin (non-metformin group). These totally 1024 patients included 513 males and 511 females, the mean age was (60.25±11.83) years (25–88 years), the duration after diabetic onset was (7.63±6.55) years (1 month–38 years).

Clinical assessment and biochemical determination
The venous whole blood samples were withdrawn from all subjects after a 10-hour overnight fast, and the urine samples for continuous 3 days were collected for the determination of albumin excretion.

The fasting plasma glucose (FPG) was assayed with glucose oxidase method. The fasting serum insulin (FINS) levels were determined with a radioimmuno-assay (RIA) kit (Linco Company, USA). The plasma Cr, blood urea nitrogen (BUN), and alanine transferase (ALT) were measured by enzymatic procedures using an autoanalyzer (Hitachi 7600-020, automatic analyzer, Japan). Hemoglobin A1c (HbA1c) was measured by high- pressure liquid chromatography (HPLC). The normal range of plasma Cr in Chinese was 35–130 µmol/L. All intra- and inter-assay coefficient variation (CV) were less than 5%. The body mass index (BMI) was calculated with body weight in kg divided by the square of the height in m².

GFR was detected by ⁹⁹technetium isotope scan.

Measurement of fasting plasma lactic acid
The measurement of lactic acid has been described elsewhere.⁸ In brief, the blood samples of all subjects were taken after an overnight fast. All plasma lactate assays were performed with enzyme-electrode method (Biosen5030 Autocal glucose-lactate analyzer, EKF diagnostic Company, Germany), the normal range for venous plasma in Chinese was 0.68–1.86 mmol/L. The intra- and inter-assay CV of lactate were 0.78% and 1.28%, respectively. The hyperlactacidemia was diagnosed when it exceeded 2.0 mmol/L.

Statistical analysis
Results were expressed as mean ± standard deviation (SD). Difference between mean values was conducted by two-tailed Student′s t test. Variables (including plasma lactate, serum insulin, and glycosylated hemoglobin) which did not conform to a normal distribution were analyzed after logarithmic transform. Comparison of proportions was analyzed by chi-square test. Associations between variables were assessed with Spearman′s rank-order correlation coefficient analysis and multiple stepwise regression analysis. Receive operating characteristic (ROC) curve analyses were used to determine appropriate cutoff of Cr in identifying the individuals with hyperlactatemia. The optimal cutoff was obtained from the Youden index (maximum (sensitivity + specificity – 1)). Greater accuracy was selected by a larger Youden index. SPSS version 13.0 (SPSS Inc., Chicago, IL, USA) was used for all data analyses. P <0.05 was considered statistically significant.

Clinical characteristics of studied subjects
Clinical characteristics of patients (n=1024) in the study were showed in Table 1. The patients in metformin- treated group had significantly greater BMI (P=0.000) and GFR (P=0.001), higher FPG (P=0.01) levels, smaller age (P=0.000), shorter diabetic duration (P=0.02), and lower FINS levels (P=0.018) than those in non-metformin group. There were no significant differences in gender, HbA1c, Cr, BUN, and ALT between two groups with or without metformin (P >0.05).

view in a new window

Table 1. Comparison of the clinical characteristics and biochemical indexes between the two groups

Divided by every 10 years of patient age, the lactate concentrations of diabetic patients were slightly higher than healthy controls in the same age-phase, but the difference was not significant, and there was no significant difference between the subgroups of different age in diabetic patients (all P >0.05, data not shown).

Factors influencing the lactic acid levels
In total subjects, the Spearman correlation analysis showed that fasting lactic acid level was positively correlated with Cr (r=0.279, P <0.001), BUN (r=0.091, P=0.004), ALT (r=0.161, P <0.001), BMI (r=0.155, P <0.001), urine albumin (Ualb) (r=0.092, P=0.003), and metformin (r=0.184, P <0.001, Table 2). It had no significant correlation with age, diabetes duration, FPG, FINS and HbA1c. After adjusted for ALT and BMI, the positive correlation between lactate and Cr still existed significantly (r=0.291, P=0.000). The stepwise multiple regression analysis indicated that factors influencing lactate acid in turn were Cr (P <0.001), metformin (P <0.001), ALT (P <0.001), BMI (P=0.001), urine albumin (Ualb) (P=0.010), and BUN (P=0.014) in all patients (Table 3).

view in a new window

Table 2. Spearman correlation-analysis about the associating factors of lactic acid

view in a new window

Table 3. Multiple stepwise regression analysis on the impacting factors of blood lactate in totally patients

In non-metformin group, the fasting lactic acid was positively associated with Cr (r=0.255, P <0.001), BUN (r=0.137, P <0.001), ALT (r=0.133, P <0.001), Ualb (r=0.111, P=0.007), and BMI (r=0.102, P=0.012, Table 2). The stepwise multiple regression analysis showed that Cr was the first entered variable that correlated with lactate acid (P <0.0001), the sequential variables were ALT (P=0.001), BMI (P=0.006), and BUN (P=0.036, Table 4).

view in a new window

Table 4. Multiple stepwise regression analysis about the factors influencing lactic acid in non-metformin group

In metformin-treated group, Cr (r=0.219, P=0.001) and ALT (r=0.136, P=0.005) had positive association with lactic acid, diabetic duration was negatively associated with it (r=–0.111, P=0.023), and the correlation of BMI with lactate was weakly but significant (r=0.096, P=0.049) (Table 2). Factors like age, diabetic duration, and FPG, had no correlation with it. The multiple stepwise regression demonstrated that Cr (P <0.001) and ALT (P=0.037) were related factors of lactic acid.

view in a new window

Figure 1. Gradual changes of lactic acid concentration in different Cr subgroups. Lactic acid level increased gradually with the elevating levels of Cr and reached its peak at Cr 111–130 µmol/L. *P <0.01 vs four subgroups with Cr ≤90 µmol/L. There was no significant difference of lactic acid levels between Cr 91–110 µmol/L subgroup and Cr 111–130 µmol/L subgroup. Error bars: 2 SE.

Furthermore, the change of plasma lactic acid levels among three subgroups of GFR, ≤60 ml∙min^–1∙1.73 m^–2 group (n=61), 61–90 ml∙min^–1∙1.73 m^–2 group (n=432), ≥91 ml∙min^–1∙1.73 m^–2 group (n=531) in total patients were compared. The results revealed that the mean lactate

concentration of ≤60 ml∙min^–1∙1.73 m^–2 subgroup ((1.13±0.47) mmol/L) had no significant difference compared with that of 61–90 ml∙min^–1∙1.73 m^–2 subgroup ((1.23±0.50) mmol/L, P=0.13) and ≥91 ml∙min^–1∙1.73 m^–2 subgroup ((1.21±0.49) mmol/L, P=0.21). Similarly, no any statistically significant difference of lactate levels was found between 61–90 ml∙min^–1∙1.73 m^–2 and ≥91 ml∙min^–1∙1.73 m^–2 subgroups (P=0.56). In non-metformin group, the mean lactate levels of three GFR subgroups that was GFR ≤60 ml∙min^–1∙1.73 m^–2 (n=30, (1.03±0.37) mmol/L), 61–90 ml∙min^–1∙1.73 m^–2 (n=272, (1.18±0.47) mmol/L), ≥91 ml∙min^–1∙1.73 m^–2 (n=297, (1.11±0.43) mmol/L) had no obvious difference (all P >0.05) either.

Optimal cutoff of Cr for hyperlactacidemia
The ROC curve of Cr and lactic acid in non-metformin group was conducted and shown in Figure 2. The area under curve (AUC) of Cr on lactic acid was 0.652 (95% CI: 0.576–0.785) and with significance (P=0.013). The optimal cutoff of Cr calculated from the Youden index was 96.5 µmol/L, and at this level the Youden index was 0.373. The sensitivity and specialty of Cr for predicting lactatemia is 61.3% and 92.6%, respectively.

view in a new window

Figure 2. Receiver operating characteristic (ROC) curve for creatinine (Cr) to detect hyperlactacidemia for type 2 diabetes without metformin treatment.

In total group, the AUC of Cr on lactate was 0.738 (95% CI: 0.654–0.902, P=0.001), and the optimal cutoff of Cr was 95.85 µmol/L, and at this level the Youden index was 0.384. The sensitivity and specialty of Cr for predicting lactatemia is 64.5% and 92.9%, respectively.

In metformin-treated group, the AUC of Cr on lactate was 0.697 (95% CI: 0.576–0.818, P=0.003), and the optimal cutoff of Cr was 95.85 µmol/L, and at this level the Youden index was 0.353. The sensitivity and specificity of Cr for predicting lactatemia is 69.2% and 94.5%, respectively.

Lactic acid was a product of anoxic metabolism in humans. Lactic acidosis in diabetes was a serious acute complication and caused almost 100% death ten years ago. Even with the great advancement of rescue technique with hemodialysis today, the mortality of lactate acidosis was almost 50%. Therefore, it is important to monitor the change of plasma lactic acid and predict the lactic acidosis on time.

It had been reported that metformin could induce hyperlactic academia, even cause fatal lactic acidosis in diabetic patients with renal dysfunction and other one or more contraindications.⁹ We showed in the present study that the mean lactic acid concentrations increased in metformin-treated group, and 9.18% patients received metformin (1.5 g/d) therapy had lactatemia, metformin administration was associated with fasting lactic acid (P <0.05). These results were concordant to previous related reports.^4,6,7,10,11 The research of Abbasi et al⁶ about 110 patients with normal renal function indicated that no matter taken alone or combined with other glucose-lowering agents, metformin may cause sub-clinical elevation of lactic acid concentrations, especially in patients with hepatic or renal dysfunction, or with low blood perfusion and oxygen deficiency condition, such as CHF and COPD. Davis et al¹⁰ also found that metformin-treated type 2 diabetic patients had higher plasma lactate concentrations than non-metformin patients, and metformin use, but not the dosage, was independently related to blood lactic acid concentrations. A study in 97 Singapore metformin-treated type 2 diabetics showed that 20.6% (20/97) had lactatemia identified by lactate levels higher than 2.2 mmol/L, however, in the subgroups with daily dosage of metformin ≤1.0 g, 1.1–2.0 g, and >2.0 g, the mean lactate concentration was 1.7, 1.6, and 2.1 mmol/L, respectively, and without significance (P=0.119), so they considered there were no correlation between total daily dose of merformin with fasting lactate level.¹¹ It was well known that lactic acidosis induced by metformin was rare in patients with normal kidney function,^4,5,9 even in elderly type 2 diabetic patients.¹² Nevertheless, on the basis of the current data, advanced age per se, mild renal impairment and stable heart failure can no longer be upheld as contraindications to the use of metformin, and the indications of metformin use included patients with normal renal function or slightly renal dysfunction.^3-6 The meta-analysis of Salpeter et al,³ reports of Rachmani et al⁴ and Papanas et al¹³ proved that diabetic patients who were treated with metformin and who tolerated the drug well may continue taking it, even when mild renal impairment develops, possibly up to serum creatinine levels of 220 µmol/L. With regards to this aspect, though lactic acidemia happened to partial diabetic patients with normal renal function, no lactic acidosis was found in present study. Thus, it was relatively safe for type 2 diabetic patients with normal plasma Cr levels to receive metformin therapy. However, the GFR of partial patients (5.96%, 61/1024) were lower than 60 ml/min despite of their Cr in normal range, the incidence of lactic academia was higher obviously in patients with Cr ≥91 µmol/L than in ≤90 µmol/L, and most of lactatemia happened in patients with Cr above 90 µmol/L. Just in case, because of the existence of sub-clinical lactatemia induced by metformin, it was necessary to measure the lactic acid concentrations in diabetic patients during metformin therapy even though their renal functions were “normal”, especially for those with Cr levels higher than 100 µmol/L, to distinguish the potentially lactic acid accumulation.

The present study showed additionally that the mean lactic acid level in type 2 diabetic patients with normal renal function and without metformin treatment was similar to that in healthy adults reported by Lu et al⁸ of our institute (P >0.05). This suggested that there was usually no metabolic abnormality of lactic acid in vivo of type 2 diabetic patients with normal renal excretion function. Divided by every 10 years of patient age, the lactate concentrations of diabetic patients were slightly higher than healthy controls in the same age-phase, but the difference had no statistical significance, and there was no significant difference between the subgroup of different age in diabetic patients (data not shown), these results were similar to the results of elderly Chinese diabetic population with normal renal function.¹¹ But on the other hand, the results of the study also indicated lactatemia existed in part of these patients (3.84%) though their renal functions were almost “normal” and they did not take metformin. These suggest that lactatemia was not only ascribed to the use of metformin, also the other possible reasons such as hypoxia and insulin resistance in obese patients.¹⁴

Simple correlation and multiple stepwise regression analysis of present study revealed that Cr, ALT, and BMI were associated factors of lactate levels despite of the administration of metformin. Among the affecting factors of lactic acid, Cr had the strongest association with it, and lactate concentrations increased gradually with the uplifting of Cr levels. Furthermore, Cr level of all 62 cases with lactatemia in this study surpassed 90 µmol/L, whatever administrated metformin. The ROC of Cr analysis demonstrated that the lactic acid levels inclined to increase when Cr level was higher than 96.5 µmol/L. The similar results came from the study reported by Rachmani et al⁴ in which lactic acid was measured in 393 type 2 diabetes with coronary heart disease (CHD), CHF or COPD, their data showed that the lactate concentration was mainly associated positively with plasma Cr levels and BMI. On the other hand, the study of Gudmundsdottir et al¹⁵ showed that the incidence of metformin-associated lactic acidosis in Norway became more frequent due to the combined use of metformin and drugs blocking the renin-angiotensin system, because the latter reduced the glomerular filtration rate, proved by the high lactate levels accompanied by high creatinine concentrations. Since the elimination and excretion of lactate was fulfilled in kidney, the lactic acidemia could be explained by the reduction of renal excretion in the condition of slight glomerular impairment. Thus it was important to detect the plasma lactic acid and to monitor the change of its concentrations for diabetic patients with low-grade renal impairment indicated by Cr level overtopping 100 µmol/L.

On the contrary, the present study failed to get the strong correlation between GFR and blood lactate levels. The comparison of lactate levels in different GFR subgroups proved that there were no significant difference of lactic acid concentration in patients with GFR ≤60 ml/min, 61–90 ml/min, or ≥91 ml/min. We were confused by these results. However, the similar results were searched by other reports, such as Khan et al,⁷ Davis et al,¹⁰ and Lim et al.¹¹ The results of Lim et al¹¹ also showed that the mean lactate levels of subgroups with ≤60 ml/min ((1.7±0.3) mmol/L), 61–90 ml/min ((1.8±0.3) mmol/L), ≥91 ml/min ((1.8±0.4) mmol/L) had no statistical significant distinction in type 2 diabetic population.¹⁰ The spearman correlation and multiple stepwise regression analysis indicated that GFR was not the independent associated factor of lactate both in diabetics treated with metformin and diabetics without metformin therapy. The possible reason of these results may be related to the imprecision of GFR detected by isotope scan ascribing to the impact of instable blood glucose at that time, because high plasma glucose could induce higher glomerular blood flow and filtration, which caused the exaggerative GFR. On the other hand, the beingless association of GFR and lactate may be explained by all of diabetics recruited in this study having normal plasma creatinine level and most of them having GFR higher than 60 ml∙min^–1∙1.73 m^–2, only GFR of 61 cases (5.96%) was lower than 60 ml∙min^–1∙1.73 m^–2., and this small sample size may limit the accuracy of mean lactate level.

Most of patients with hyperlactatemia were asymptomatic; a few of them felt fatigue, or leg sore, and other non-specific symptoms. The 62 cases of present study with higher lactic acid levels did not have differential symptoms, either. Thus, it was essential to find a clinical marker to predict the lactic acidemia and to prevent the lactic acidosis. Though Cr level was an insensitive index to reflect the renal function，more than 70% glomeruli were damaged when Cr was higher than 130 µmol/L, the excretion of lactate likely decreased and the lactatemia appeared when the partial glomerular function lost indicating by slight elevating Cr levels (over 100 µmol/L). Researchers had found diabetic patients with high lactate levels had higher blood anion gap (AG), which predicted the lactatemia.⁷ But their study could not prove that AG was better than measuring fasting lactic acid directly. On the contrary, results of present study indicated that Cr has positive relevance with the lactate concentrations, it was an independent associated factor of lactic acid in spite of in total patients, non-metformin group, or metformin- treated group, and its AUC in ROC for evaluating lactic acid got the significant 0.652, 0.738, and 0.697 in non-metformin group, total group, and metformin-treated group, respectively. We further revealed the optimal cutoff of Cr was nearly the same level of 96.5 µmol/L, 95.85 µmol/L and 95.85 µmol/L in above turn. Therefore, Cr can be used to forecast the risk of lactic acidemia, and the optimal cutoff of it to induce lactic acidemia was 96.5 µmol/L suggested by ROC analysis in the study. The specificity of Cr at this level to predict the lactatemia exceeded 90%. Compared with the complex and expensive GFR measurement, the easily detected and practical Cr assay might be a better alternative to discriminate the lactatemia after metformin administration.

Another interesting result of this study is the positive association of BMI with lactic acid level, and the similar outcome had been reported by Rachmani et al,⁴ Abbasi et al,⁶ Davis et al,¹⁰ and Khan et al.⁷ This may imply that the plasma lactic acid levels tend to elevate with the increasing of body fat. Recently, Qvisth et al¹⁴ revealed that both skeletal muscle and adipose tissue are significant sources of lactate production in the post-absorptive state, and insulin stimulates net lactate release in both tissues, but in obese women with insulin-resistance, the ability to increase lactate release in response to insulin is impaired ascribing to their defective insulin regulation of both tissue lactate metabolism and local blood flow. Results reported by Sandqvist et al¹⁶ also found that there was an enhanced net lactate release per fat cell in abdominal subcutaneous tissue of first-degree relatives of type 2 diabetic patients.¹⁵ Crawford et al¹⁷ onwards found that the median lactate levels were significantly higher in both the obesity subjects and obese subjects with metabolic syndrome compared with lean controls. After 12–20-week very low-calorie diet intervention, the lactate levels of group with both obesity and metabolic syndrome group fell significantly with the 31% decrease of BMI. Huo et al¹⁸ found increased trimethylamine-N-oxide (TMAO), 3-hydroxybutyrate (3-HB) and decreased glucose, N-acetyl glycoprotein (NAC), lipoprotein, lactate, acetoacetate and unsaturated lipids in serum from metformin treated patients; and Dykens et al¹⁹ additionally revealed that biguanide-induced lactic acidosis can be attributed to acceleration of glycolysis in response to mitochondrial impairment. Therefore, we speculate that the higher blood lactate concentrations in patients with bigger BMI may be explained by the lasting oxygen-deficient environment because of obesity and insulin resistance, this induce the increment of the production and release of lactic acid from adipose cell through activating anaerobic glycolysis pathway. So far, the investigations about the influence of obesity on the plasma lactate levels is rare, further studies need to undertake to ascertain the molecular mechanism under the relevance of visceral adipose accumulation and lactate metabolism.

In summary, we can conclude from the present study that there is low proportion of lactatemia in type 2 diabetes despite of their normal renal function and without metformin use, and metformin can increase the incidence of hyperlactic acidemia, but not fatal lactic acidosis in diabetics with Cr <130 µmol/L. The plasma levels of Cr, ALT, and BMI are important associating factors of blood lactate concentrations. The plasma Cr level is a practical index associated with lactic acid, and the optimal cutoff of Cr is 96.5 µmol/L in the prediction of lactatemia.

1. Chinese Medical Association. Hypoglycemic treatment of type 2 diabetes. Chinese guideline for the diagnosis and treatment of type 2 diabetes (2007 edition). Beijing: Publisher of Beijing University; 2008: 10-11.

2. Di Grande A, Vancheri F, Giustolisi V, Giuffrida C, Narbone G, Licata M, et al. Metformin-induced lactic acidosis in a type 2 diabetic patient with acute renal failure. Clin Ter 2008; 159: 87-89.

3. Salpeter S, Greyber E, Pasternak G, Salpeter E. Risk of fatal and nonfatal lactic acidosis with metformin use in type 2 diabetes mellitus: systematic review and meta-analysis. Cochrane Database Syst Rev 2006; 25: CD002967.

4. Rachmani R, Slavachevski I, Levi Z, Zadok B, Kedar Y, Ravid M. Metformin in patients with type 2 diabetes mellitus: reconsideration of traditional contra-indications. Eur J Intern Med 2002; 13: 428-431.

5. Holstein A, Stumvoll M. Contraindications can damage your health – is metformin a case in point? Diabetologia 2005; 48: 2454-2459.

6. Abbasi AA, Kasmikha R, Sotingeanu DG. Metformin-induced lacticacidemia in patients with type 2 diabetes mellitus. Endocr Pract 2000; 6: 442-446.

7. Khan JK, Pallaki M, Tolbert SR, Hornick TR. Lactic acidemia associated with metformin. Ann Pharmacother 2003; 37: 66-69.

8. Lu HJ, Tang JL, Yuan QY, Jia WP, Xiang KS. The normal range of plasma lactic acid in Chinese. Shanghai Med J (Chin) 2005; 28: 27-29.

9. Seidowsky A, Nseir S, Houdret N, Fourrier F. Metformin- associated lactic acidosis: a prognostic and therapeutic study. Crit Care Med 2009; Epub ahead of print.

10. Davis TM, Jackson D, Davis WA, Bruce DG, Chubb P. The relationship between metformin therapy and the fasting plasma lactate in type 2 diabetes: The Fremantle Diabetes Study. Br J Clin Pharmacol 2001; 52: 137-144.

11. Lim VC, Sum CF, Chan ES, Yeoh LY, Lee YM, Lim SC. Lactate levels in Asian patients with type 2 diabetes mellitus on metformin and its association with dose of metformin and renal function. Int J Clin Pract 2007; 61: 1829-1833.

12. Tian H, Li CL, Yang G, Wang YZ, Zong WY, Shao YH, et al. Safety of metformin in the treatment of elderly type 2 diabetes mellitus. Chin J Intern Med (Chin) 2008; 47: 914-918.

13. Papanas N, Monastiriotis C, Christakidis D, Maltezos E. Metformin and lactic acidosis in patients with type 2 diabetes – from pride and prejudice to sense and sensibility. Acta Clin Belg 2009; 64: 42-48.

14. Qvisth V, Hagström-Toft E, Moberg E, Sjöberg S, Bolinder J. Lactate release from adipose tissue and skeletal muscle in vivo: defective insulin regulation in insulin-resistant obese women. Am J Physiol Endocrinol Metab 2007; 292: E709-E714.

15. Gudmundsdottir H, Aksnes H, Heldal K, Krogh A, Froyshov S, Rudberg N, et al. Metformin and anti-hypertensive therapy with drugs blocking the renin-angiotensin system, a cause of concern? Clin Nephrol 2006; 66: 380-385.

16. Sandqvist MM, Eriksson JW, Jansson PA. Increased lactate release per fat cell in normoglycemic first-degree relatives of individuals with type 2 diabetes. Diabetes 2001; 50: 2344-2348.

17. Crawford SO, Ambrose MS, Hoogeveen RC, Brancati FL, Ballantyne CM, Young JH. Association of lactate with blood pressure before and after rapid weight loss. Am J Hypertens 2008; 21: 1337-1342.

18. Huo T, Cai S, Lu X, Sha Y, Yu M, Li F. Metabonomic study of biochemical changes in the serum of type 2 diabetes mellitus patients after the treatment of metformin hydrochloride. J Pharm Biomed Anal 2009; 49: 976-982.

19. Dykens JA, Jamieson J, Marroquin L, Nadanaciva S, Billis PA, Will Y. Biguanide-induced mitochondrial dysfunction yields increased lactate production and cytotoxicity of aerobically-poised HepG2 cells and human hepatocytes in vitro. Toxicol Appl Pharmacol 2008; 233: 203-210.