2023 Volume 70 Issue 1 Pages 121-128
Metformin monotherapy as first-line treatment for patients with type 2 diabetes (T2D) has been shown to effectively improve blood glucose levels and motivation to undergo treatment and prevent complications. However, no studies have reported its effect when combined with other drugs or compared the effect based on administration time. This study aimed to investigate the effect of metformin administration in Japanese patients with T2D, examine how the introduction line impacts the effect of metformin, and examine the characteristics of patients demonstrating improved blood glucose levels. Data on characteristics of patients who were newly prescribed metformin with no shifting of hypoglycemic agents in the subsequent 24-week observation period, and their age [mean, 56.8 years], body mass index [mean, 27.5 kg/m2], glycated hemoglobin [HbA1c] [mean, 8.1%], and duration of diabetes [mean, 3.0 years] were obtained from the medical records of 201 patients. The changes in HbA1c by introduction line after 24 weeks were –1.59%, –0.91%, –0.89%, and –0.65% in the first, second, third, and fourth induction lines, respectively; earlier introduction more significantly improved blood glucose. The factors significantly associated with HbA1c changes were early introduction, high baseline HbA1c, high estimated glomerular filtration rate, decreased insulin secretion, short estimated duration of diabetes, and increased metformin dose. Furthermore, factors contributing to the largest HbA1c improvement by metformin were high baseline HbA1c and early administration. Metformin is expected to lower blood glucose levels in Japanese patients with T2D, even in those with decreased insulin secretion, due to its early introduction as a first-line drug.
INTERNATIONAL GUIDELINES, such as the American Diabetes Association and European Association for the Study of Diabetes recommend metformin as an initial treatment for patients with type 2 diabetes (T2D) who cannot meet their glycemic goals despite diet and other lifestyle interventions [1]. Metformin is the first-line drug used in several countries outside of Japan because of the clear evidence of suppressing macroangiopathy [2, 3], its safety [4], and its comparatively low price. However, the pathological conditions, body shapes, and lifestyle of Japanese patients with T2D are different from those in other countries. Owing to the lack of research, Japanese diabetes guidelines do not have a standard first-line drug; thus, drugs for patients with T2D are selected according to individual pathological conditions [5]. Often, dipeptidyl peptidase-4 (DPP-4) inhibitors are employed as the first-line T2D drug in Japan [6, 7]. However, the price for metformin for X 1,000 mg/day is approximately one-fifth of the cost of the usual DPP-4 inhibitors dose. Reportedly, total medical costs are the lowest in patients started on biguanides, such as metformin [8]; therefore, adopting them could improve treatment outcomes as high medical costs are cited as a reason for treatment discontinuation in patients with T2D in Japan [9]. Early initiation of metformin treatment may improve blood glucose levels, maintain long-term motivation, and prevent complications in patients with T2D.
Although the effect of metformin when used alone as a first-line drug has been established, no study has investigated the effect of adding metformin to the treatment regimen of patients who were already maintained on other medications.
Evaluating the hypoglycemic effect of administering metformin in the first, second, third, fourth, and subsequent lines of treatment is difficult as metformin is generally used as the first-line of treatment. Moreover, although metformin is often used as a treatment to combat insulin resistance in patients with obesity, few studies have discussed factors predicting its efficacy in Asians.
Therefore, we aimed to investigate metformin usage status in Japanese patients, examine how the time of introduction impacts the effect of metformin, and analyze the clinical factors contributing to the metformin hypoglycemic effect.
This retrospective study was conducted at multiple institutions in accordance with the Declaration of Helsinki. It was registered with University Hospital Medical Information Network (UMIN000040854) with the approval of the Clinical Research Committee of Yokohama City University Hospital (approval number: B200600030). Obtaining individual informed consent was difficult because many patients had not visited the hospital, and some had different doctors for each consultation. Therefore, a document for information disclosure was created, where the research purpose and outline at each research institution was disclosed. Patients were provided the opportunity to refuse the use of their information.
Study participants included patients with T2D aged between 20 and 80 years who attended the Department of Endocrinology and Diabetes at Yokohama City University Medical Center and the Urafune Kanazawa Internal Medicine Clinic, were newly prescribed metformin from January 1, 2008 to November 30, 2019, and were followed-up for the next 24 weeks.
At the start of metformin administration, the administration of other diabetic drugs was not discontinued and the administered dose did not increase or decrease. Eligible patients were those who had not discontinued, changed, or added a hypoglycemic drug other than metformin within 24 weeks after metformin prescription. Patients initially administered insulin injections were included in this study; even if insulin units had been changed or discontinued during the 24-week observation period. The exclusion criteria were changes in diabetes drugs other than metformin during the 24 weeks, patients with extremely poor medication compliance based on review of medical records, steroids use, cancer under aggressive treatment, hospitalization during the observation period, and severe infections and trauma. We reviewed the medical records of the hospital and identified 1,762 patients who were prescribed metformin; of these, 201 patients met the inclusion and exclusion criteria.
Patients were classified into four groups according to the line (first, second, third or fourth) of the introduction of metformin. Cases in which metformin was introduced as the fifth agent or later were aggregated in the fourth. Patient backgrounds were compared among the categories. Patient information, clinical data, blood and urine results, and medical records were obtained 24 weeks after the start of metformin administration from the medical records. However, data on glycated hemoglobin (HbA1c) values were also obtained at 12 weeks. The HbA1c value in this study is expressed according to the National Glycohemoglobin Standardization Program value [10]. The HbA1c values from the period when the Japan Diabetes Society value had been used were converted by adding 0.4. The blood glucose level in this study was the blood glucose taken at any time, beta-cell function (HOMA2-%β), and insulin resistance (HOMA2-IR) were calculated via a homeostatic model assessment (HOMA) Calculator using plasma glucose levels and C-peptides or insulin (https://www.dtu.ox.ac.uk/homacalculator/download.php). Insulin or C-peptide levels were assessed in patients who did not receive insulin, and C-peptide levels were assessed in patients who received insulin.
Statistical analysisStatistical analyses were performed using JMP Pro 15.0.0 software (SAS Institute, Cary, NC, USA). The results of the mean and proportion of participants’ characteristics were expressed as either mean (± standard deviation), median (25–75% interquartile range), or numbers with percentages. One-way analysis of variance was performed for normally distributed data, and the nonparametric method was performed for non-normally distributed data. The four categories were compared using Dunnett’s multiple comparison method, and the results were expressed as mean (± standard error). Univariate and multivariate linear regression analyses were conducted to identify factors that were independently correlated with changes in HbA1c levels. HOMA2-%β, HOMA2-IR, and estimated duration of diabetes were transformed to natural log values to enable statistical analyses since the frequency distributions of these variables were skewed. Univariate analysis was first conducted, and variables such as the introduction line, baseline HbA1c, estimated glomerular filtration rate (eGFR), HOMA2-%β, estimated duration of diabetes, and dose of metformin at 24 weeks, which satisfied the criterion of p < 0.05, were collectively entered in the multivariate analyses model, with age, sex, and body mass index (BMI) as background variables and changes in the number of insulin units24-0W as factors that affect HbA1c levels. In addition, the standard partial regression coefficient (β) was used to compare the degree to which each explanatory variable affected the change in HbA1c level. All independent variables in the multivariate linear regression analyses were tested for multicollinearity to ensure that the variance inflation factor did not exceed 2. Statistical significance was set at p < 0.05.
The baseline characteristics of 201 patients are presented in Table 1. There were 3 cases in which metformin was introduced as the fifth agent or later. Specifically, in 2 cases, it was introduced as the fifth agent and in 1 case as the sixth, and these aggregated as the fourth agent. Only diabetes duration (years) was different between the groups, 2.0 as first-line medication, 3.5 as second, 8.0 as third, and 9.5 as fourth line, and we found that the longer the diabetes duration, the later the time of metformin introduction (Table 2). The baseline HbA1c (8%) was not significantly different among the groups, and BMI, renal function, HOMA2-%β, and HOMA2-IR were not significantly different (Table 2). The overall average change in HbA1c after 24 weeks was –1.1 ± 0.1% and –1.59 ± 0.2% as first-line medication, –0.91 ± 0.1% as second (p < 0.05, vs. first-line), –0.89 ± 0.2% (p < 0.05, vs. first-line) as third, and –0.65 ± 0.2% (p < 0.05, vs. first-line) as fourth line (Table 3, Fig. 1). The earlier the introduction of metformin, the greater the rate of decreased HbA1c, and the greater the achievement of target of HbA1c ≤7% (Table 3). The mean of urine albumin levels was 18.8 [8.1, 42.8] mg/g Cr (n = 161) at baseline and 19.6 [8.2, 48.4] mg/g Cr (n = 147) at week 24, which was not significant.
| Age (years) | 56.8 ± 11.1 |
| Sex (male/female) | 137/64 |
| Weight (kg) | 75.4 ± 17.0 |
| BMI (kg/m2) | 27.5 ± 5.4 |
| Estimated duration of diabetes (years) | 3.0 (2.0, 8.0) |
| Blood glucose (mg/dL) | 177.2 ± 59.2 |
| HbA1c (%) | 8.1 ± 1.4 |
| AST (U/L) | 19 ± 6.5 |
| ALT (U/L) | 18 ± 9.2 |
| γ-GTP (U/L) | 28.5 ± 8.4 |
| eGFR (mL/min/1.73 m2) | 82.2 ± 20.0 |
| HOMA2-%β | 61.8 (36.7, 102.3) |
| HOMA2-IR | 2.2 (1.3, 3.3) |
| Hypoglycemic treatment | |
| -No medication (n, (%)) | 76 (38) |
| -OHA only (n, (%)) | 83 (41) |
| -GLP-1 ± OHA (n, (%)) | 4 (2) |
| -Insulin ± GLP-1 ± OHA (n, (%)) | 37 (19) |
| Neuropathy (%) | 17.3 |
| Nephropathy (%) | 34.5 |
| Retinopathy (%) | 23.3 |
| Macroangiopathy (%) | 5.8 |
| Hypertension (%) | 45.3 |
| Dyslipidemia (%) | 70.9 |
Values are expressed as mean ± standard deviation, median (25–75% interquartile range) or percentage (%).
BMI, body mass index; HbA1c, glycated hemoglobin; AST, aspartate transaminase; ALT, alanine transaminase; γ-GTP, γ-glutamyl transpeptidase; eGFR, estimated glomerular filtration rate; HOMA, homeostatic model assessment; OHA, oral hypoglycemic agent; GLP-1, glucagon-like peptide-1
Values of HOMA2-%β and HOMA2-IR were calculated using the HOMA Calculator with plasma glucose levels and C-peptides or insulin (https://www.dtu.ox.ac.uk/homacalculator/download.php). The patient characteristics were defined as follows: neuropathy, the presence of a subjective symptom (palsy, pain, etc.) or the weakening of the Achilles tendon reflex, sense of vibration, or sensory nerve conduction velocity; nephropathy, diabetic nephropathy stages 2, 3, 4, or 5; retinopathy: simple, preproliferative, or proliferative diabetic retinopathy; macroangiopathy: ischemic heart disease, cerebral infarction, and arteriosclerosis obliterans.
| Introduction line | First line | Second line | Third line | Fourth line | p value |
|---|---|---|---|---|---|
| Subjects (n, (%)) | 77 (38) | 63 (31) | 39 (20) | 22 (11) | — |
| Age (years) | 55.1 ± 10.5 | 56.6 ± 10.7 | 59.2 ± 11.6 | 58.9 ± 13.4 | 0.227 |
| Sex (male/female) | 50/27 | 43/20 | 28/11 | 16/6 | 0.848 |
| Weight (kg) | 76.8 ± 14.0 | 74.7 ± 18.0 | 74.1 ± 21.7 | 75.1 ± 15.2 | 0.829 |
| BMI (kg/m2) | 28.1 ± 4.5 | 27.3 ± 6.0 | 27.0 ± 6.5 | 27.1 ± 4.7 | 0.703 |
| Estimated duration of diabetes (years) | 2.0 (0.9, 3.0) | 3.5 (2.0, 7.1) | 8.0 (4.0, 17.0) | 9.5 (4.8, 13.5) | <0.0001 |
| Blood glucose (mg/dL) | 177.2 ± 51.4 | 165.3 ± 61.2 | 192.8 ± 69.2 | 183.4 ± 56.4 | 0.139 |
| HbA1c (%) | 8.3 ± 1.7 | 7.8 ± 1.1 | 8.3 ± 1.4 | 8.2 ± 0.9 | 0.203 |
| eGFR (mL/min/1.73 m2) | 84.6 ± 20.8 | 81.8 ± 15.9 | 78.9 ± 23.5 | 80.7 ± 20.6 | 0.509 |
| HOMA2-%β | 70.0 (40.9, 112.3) | 58.5 (35.9, 85.4) | 70.3 (28.2, 108.3) | 55.6 (29.3, 91.8) | 0.255 |
| HOMA2-IR | 2.3 (1.3, 3.2) | 2.3 (0.9, 2.9) | 2.3 (1.6, 4.2) | 1.9 (1.5, 2.8) | 0.063 |
p, probability; BMI, body mass index; HbA1c, glycated hemoglobin; eGFR, estimated glomerular filtration rate; HOMA, homeostatic model assessment
One-way ANOVA and nonparametric method were performed, the results are expressed as mean ± standard error, median (25–75% interquartile range) or percentage.
Values of HOMA2-%β and HOMA2-IR were calculated using the HOMA Calculator with plasma glucose levels and C-peptides or insulin (https://www.dtu.ox.ac.uk/homacalculator/download.php).
| Introduction line | Total | First line | Second line | Third line | Fourth line | p value |
|---|---|---|---|---|---|---|
| Dose of metformin (mg/day) | 829.6 ± 34.9 | 1,013.0 ± 54.0 | 662.7 ± 59.7 | 756.4 ± 75.9 | 795.5 ± 101.1 | 0.0002 |
| Change in weight (kg) | –0.6 ± 0.2 | –1.6 ± 0.4 | 0.02 ± 0.5 | 0.3 ± 0.3 | –0.2 ± 0.6 | 0.013 |
| Change in weight (%) | –0.7 ± 0.3 | –1.9 ± 0.5 | 0.2 ± 0.8 | 0.3 ± 0.4 | –0.4 ± 0.8 | 0.032 |
| Change in HbA1c (%) | –1.1 ± 0.1 | –1.6 ± 0.2 | –0.9 ± 0.1 | –0.9 ± 0.2 | –0.6 ± 0.2 | 0.006 |
| HbA1c ≤7% achievement (%) | 62.7 | 76.6 | 67.7 | 43.6 | 36.3 | 0.0002 |
One-way ANOVA was performed, and the results were expressed as mean ± standard error or percentage.
p, probability; HbA1c, glycated hemoglobin
Since the 24-week weight of 17 subjects could not be obtained from the medical records, the baseline weight was substituted for the 24-week weight.
Change in HbA1c after 24 weeks by metformin introduction line
Patients were classified into four groups according to the line (1st, 2nd, 3rd, and 4th) of the introduction of metformin. Cases in which metformin was introduced as the 5th agent or later were aggregated in the 4th. Comparisons among the four categories were performed using Dunnett’s multiple comparison test (control = first). The results are expressed as means. * p < 0.05 vs. first at 12 weeks, ** vs. first at 24 weeks.
The average body weight change and metformin dose were –0.6 ± 0.2 kg, and 829.6 ± 34.9 mg, respectively (Table 3). Simple regression analysis showed that the factors significantly associated with HbA1c decline 24 weeks after the start of metformin treatment were early introduction line, increase in baseline HbA1c, high eGFR, a decline in HOMA2-%β, and short estimated diabetes duration. However, the relationship between the baseline BMI and HOMA2-IR was not significant (Table 4). Multiple regression analysis revealed that the strongest determinants of decreased HbA1c of metformin were the early introduction line and high baseline HbA1c (Table 4).
| Univariate | Multivariate | |||||
|---|---|---|---|---|---|---|
| B | 95% CI | p value | β | 95% CI | p value | |
| Introduction line (first–fourth) | 0.32 | (0.13, 0.52) | 0.002 | 0.28 | (0.13, 0.42) | <0.0001 |
| Baseline HbA1c | –0.82 | (–0.91, –0.74) | <0.0001 | –1.17 | (–1.30, –1.03) | <0.0001 |
| Age | 0.01 | (–0.72 × 10–2, 0.03) | 0.234 | –0.07 | (–0.21, 0.08) | 0.381 |
| Sex | 0.21 | (–0.23, 0.65) | 0.339 | –0.02 | (–0.14, 0.10) | 0.780 |
| BMI | –0.21 × 10–2 | (–0.04, 0.04) | 0.911 | 0.11 | (–0.01, 0.24) | 0.079 |
| eGFR | –0.02 | (–0.03, –0.01) | <0.0001 | –0.09 | (–0.24, 0.05) | 0.199 |
| log (HOMA2-%β) | 0.39 | (0.13, 0.65) | 0.003 | –0.02 | (–0.14, 0.11) | 0.780 |
| log (HOMA2-IR) | –0.19 | (–0.60, 0.22) | 0.360 | |||
| log (estimated duration of diabetes) | 0.29 | (0.06, 0.51) | 0.015 | 0.06 | (–0.08, 0.21) | 0.373 |
| Dose of metformin at 24 weeks | –0.1 × 10–2 | (–0.15 × 10–2, –0.74 × 10–3) | <0.0001 | –0.02 | (–0.15, 0.11) | 0.751 |
| Change in the number of insulin units24-0W | –0.03 | (–0.09, 0.03) | 0.292 | 0.06 | (–0.06, 0.18) | 0.348 |
Sex: male = 1, female = 0.
B, partial regression coefficient; β, standardized partial regression coefficient; 95% CI, 95% confidence interval; p, probability; HbA1c, glycated hemoglobin; BMI, body mass index; eGFR, estimated glomerular filtration rate; HOMA, homeostatic model assessment
Values of HOMA2-%β and HOMA2-IR were calculated using the HOMA Calculator with plasma glucose levels and C-peptides or insulin (https://www.dtu.ox.ac.uk/homacalculator/download.php).
The participants were divided into four groups according to the order of metformin introduction, as first, second, third, and fourth line of medication. The hypoglycemic effects among the categories were compared, and the clinical factors that contributed to the hypoglycemic effect of metformin were identified. Early introduction of metformin significantly decreased the HbA1c levels, and its hypoglycemic effect was observed even in patients with decreased insulin secretion, regardless of their BMI. This study showed that early introduction and high baseline HbA1c are important factors for the successful treatment with metformin. Early introduction of metformin could therefore be effective in a wide range of patients with T2D in Japan.
Actual situation of administration of metformin in patients with type 2 diabetesMetformin was started at a BMI of 27 kg/m2 and baseline HbA1c of approximately 8% at any introduction line, and there was no significant difference between groups. Previous studies in Japan reported that metformin use was associated with younger age, shorter diabetes duration, obesity, and HbA1c <8% [11]. This drug, positioned as an insulin resistance-improving drug, tends to be used in obese patients with short illnesses. Although, the approved dose of metformin in Japan is 2,250 mg/day, the average metformin dose at 24 weeks was approximately 800 mg/day, and it was higher in the group introduced on the first line. When used as the first line, the dose increases and gradually leads to treatment with a single agent. There are two possible reasons for the lower dose used in this study. First, only 750 mg/day was approved in Japan until May 2010; second, a small dose may have achieved the therapeutic target. The HbA1c reduction in this study was –1.6% on the first-line and –0.9% on the second and third lines. This agrees with the HbA1c improvement observed in overseas studies where a high metformin dose of 1,000–2,000 mg was administered as the first-line [12-16]. It is widely recognized that patients with T2D in East Asians are less obese, have a lower β-cell function, and are more sensitive to insulin than Caucasians [17-19]. Additionally, Japanese patients have impaired suppression of endogenous glucose production, which is particularly prominent in lean individuals [20]. Even a small metformin dose suppresses hepatic glucose production and exerts glycemic effects [21]. Thus, these pathophysiological differences affect the responsiveness to the drug. A sufficient HbA1c lowering effect can be expected even with a smaller metformin dose in Japan than in other countries.
Effect of metformin in patients with decreased insulin secretory capacityThis study showed that early metformin introduction, high baseline HbA1c, decreased insulin secretory capacity, high eGFR, an increased metformin dose, and short estimated diabetes duration effectively reduced HbA1c levels. A Japanese study showed that metformin further decreased HbA1c levels in patients with reduced β-cell function [22], which is consistent with the results of this study. A systematic review of studies reported that metformin administration to patients with type 1 diabetes results in a decrease in HbA1c of about 0.6–0.9% [23]; even participants with depleted endogenous insulin secretion have this degree of blood glucose improving effect. The reason for the good response of subjects with decreased β-cell function to metformin may be mediated by insulin-independent action suppression of gluconeogenesis [24-28], improvement of insulin resistance [29], activation of the intestinal-brain-liver circuit [30], changes in the intestinal flora [31, 32], and excretion of sugar in the stool [33]. However, the degree to which this lowers blood glucose is unclear.
There was no significant relationship between the rate of decrease in HbA1c and baseline BMI or HOMA2-IR indicating insulin resistance, which was similar to that reported in Asian studies [22, 34-36]. This drug is widely used in patients with T2D, who are mainly obese and insulin-resistant. However, in reality, it exerts its full effect in cases of decreased insulin secretion, regardless of the presence or absence of obesity.
We additionally investigated the clinical factors that predicted the potential efficacy of metformin in patients. Multiple regression analysis showed that early introduction and high baseline HbA1c were factors that independently improved blood glucose with metformin. During the 24-week observation period, the mean change in the number of insulin units in insulin-administered patients was –0.51 ± 7.9 U/day. Because the number of insulin units affect HbA1c, changes in the number of insulin units 24-0W, which is the number of insulin units used at 24 weeks minus that used at baseline, was added as a factor in the multiple regression analysis; however, the HbA1c results were not affected.
From these results, it is considered that even for Japanese people, who tend to be lean and have low insulin secretion, the early introduction of metformin may be expected to lower blood glucose effectively.
Delayed introduction diminishes the hypoglycemic effect of metforminIt is generally known that the longer the duration of diabetes, the lower the pancreatic β-cell function [37, 38]. There was no significant difference between the groups in the baseline HOMA2-%β value and insulin secretory capacity in the target population of this study. Therefore, the poor HbA1c improvement in patients with late metformin introduction may be due to factors other than insulin secretory capacity. First, patients with late introduction also had a significantly longer estimated duration of diabetes and may have reduced motivation for treatment, such as reduced medication adherence or decreased compliance with diet and exercise therapy. It has been reported that the longer the duration of diabetes, the more difficult it is to obtain the effect of nutritional guidance, which is one of the self-care behaviors of these patients [39]. Second, the first-line of introduction includes patients who started drug therapy shortly after their diagnosis with diabetes, and the effect of starting diet and exercise therapy at the same time could have contributed to the decrease in HbA1c. The first line group showed the greatest weight loss at 24 weeks after the initiation of metformin, suggesting that the non-pharmacologic therapy, such as patient self-care behavior, may have also contributed to the HbA1c decline in the first-line group. Metformin has a synergistic effect with other hypoglycemic agents; a well-known combination of DPP-4 inhibitors and metformin has been reported to enhance glucagon-like peptide-1 (GLP-1) action [40]. However, metformin has greater benefit for hypoglycemia when introduced early as first-line treatment than when used later as an adjunct due to drug interactions.
Can metformin be the first line of Japanese type 2 diabetes treatment?Reportedly, any hypoglycemic drug can be expected to be more effective with early introduction; however, it is unclear which drug is most effective because it has not been positively investigated head-to-head. A meta-analysis study reported that metformin and repaglinide were the most effective monotherapy options for the first line of treatment among the hypoglycemic agents [41]. We believe that we could provide the following useful information: high HbA1c levels and early introduction of metformin are key to successful treatment to continue using this drug in Japanese T2D patients; these factors are more important than dose, duration of diabetes, BMI, and pancreatic β-cell function. This study clarified the types of patients for whom metformin should be used as a first-line drug, and we believe that this drug is effective in a wide range of patient groups regardless of body size or pancreatic β-cell function. There is a need for clinical trials that can compare metformin with other diabetic drugs in assessing it use as a first-line drug for Japanese patients with T2D.
LimitationsThis study had some limitations. First, the evaluation of insulin secretory capacity and resistance is a reference value because blood was collected at any time rather than at a standardized time on an empty stomach; however, we believe that evaluation under fasting conditions is desirable. Second, medication compliance was unclear because this study was based only on medical records. Third, the actual usage of metformin is limited because the target facilities of this study were the Department of Diabetes in a university hospital and a clinic specializing in lifestyle-related diseases, including diabetes. Forth, the effect of metformin reported in this study may not have been caused by only metformin because we did not assess the use of or increase the dose of other diabetic drugs other than insulin during the observation period; therefore, cases in which HbA1c was not improved after the administration of metformin were inevitably excluded.
ConclusionThis study showed that metformin could be expected to improve blood glucose levels even in patients with decreased insulin secretion, regardless of their physique; it can be used for a wide range of Japanese T2D patients. We also showed that early metformin administration presented the highest efficiency in reducing the high HbA1c levels.
In the future, conducting clinical trials by comparing multiple hypoglycemic agents as a first-line drug, including metformin, may provide guidelines for treatment options for Japanese T2D patients.
The authors have declared that there are no competing interests.
HbA1c, glycated hemoglobin; DPP-4, dipeptidyl peptidase-4; HOMA, homeostatic model assessment; eGFR, estimated glomerular filtration rate; BMI, body mass index; GLP-1, glucagon-like peptide-1; AST, aspartate transaminase; ALT, aspartate transaminase; γ-GTP, γ-glutamyl transpeptidase; OHA, oral hypoglycemic agent
Toshihiro Misumi offered statistical guidance and Editorial support in the form of medical writing was provided by Editage (www.editage.com)
