- Email: [email protected]
Should physicians prescribe metformin to women with polycystic ovary syndrome PCOS?
Annales d’Endocrinologie 71 (2010) 25–27
Consensus of French Endocrine Society on female hyperandrogenism
Should physicians prescribe metformin to women with polycystic ovary syndrome PCOS? Rôle de la metformine dans le SOPK L. Duranteau a,∗ , P. Lefevre b , N. Jeandidier c , T. Simon d , S. Christin-Maitre e a
Service d’endocrinologie pédiatrique, hôpital Saint-Vincent-de-Paul, 82, avenue Denfert-Rochereau, 75014 Paris, France b Service d’endocrinologie, hôpital Lapeyronie, 34925 Montpellier, France c Service d’endocrinologie, Hospices civils, 67091 Strasbourg, France d Service de pharmacologie, Urcest, hôpital Saint-Antoine, 75571 Paris, France e Service d’endocrinologie, hôpital Saint-Antoine, Paris, France Available online 15 January 2010
Abstract 1. Metformin is not efficient enough in order to regulate menstrual cycles. 2.Metformin is not efficient enough in order to treat hyperandrogenism. 3. Metformin should not be used as a first-line treatment in order to treat infertility. Clomiphene citrate (CC) is the reference treatment. 4. Metformin in addition to CC is not recommended as a second line treatment, after the failure of CC alone. 5. Metformin should not be used during pregnancy in non diabetic women with PCOS, in order to prevent the risk of gestational diabetes. 6. Metformin should be prescribed to PCOS women when they are diabetic, in order to prevent their cardiovascular risk, after lifestyle modification. 7. Metformin should not be used in PCOS non diabetic women in order to lose weight. Metformin should not be used in order to treat dyslipidemia in women with PCOS. 8. In PCOS women, without diabetes, but with fasting hyperglycemia or carbohydrate intolerance, metformin should be prescribed if: BMI > 35. © 2009 Published by Elsevier Masson SAS. Keywords: Metformin; Polycystic ovary syndrome; French consensus
Metformin has been used in Europe since the seventies in order to treat type 2 diabetes. It has been prescribed in the United States since 1994. This molecule improves insulin resistance as it increases glucose capture in muscles and decreases neoglucogenesis in the liver. Women with polycystic ovary syndrome (PCOS) present insulin resistance with hyperinsulinemia in 50 to 70% of cases [1]. Therefore, in theory, metformin represents a good candidate to treat PCOS. The use of metformin has been studied in order to regulate cycle disturbances. It regulates menstrual frequency, using doses ranging from 1000 to 2500 mg per day. However, only few randomized studies have been performed and the duration of treatment reached a maximum of one year. Metformin’s efficacy remains lower than combined oral contraceptive pill [2].
∗
DOI of original article:10.1016/j.ando.2009.12.006. Corresponding author. E-mail address: [email protected] (L. Duranteau).
0003-4266/$ – see front matter © 2009 Published by Elsevier Masson SAS. doi:10.1016/j.ando.2009.12.005
Progestins alone remain the treatment of reference in order to regulate menstrual bleeding (Table 1). Metformin has been used in order to reduce hyperandrogenism and hirsutism. According to different meta-analysis, its effect is not higher than placebo and lower than antiandrogen treatments [3] (Table 2). In order to treat infertility, many hopes arrived in the late nineties, concerning metformin treatment. However, most recent randomized studies performed on large groups of patients have demonstrated that it induces a lower number of live births than clomiphene citrate (CC) [4]. Therefore, CC should remain the first line treatment in order to induce ovulation and pregnancy in women with PCOS, after lifestyle modifications (Table 3). In CC resistant women, metformin has been added to CC. This treatment has a low cost and induces few ovarian hyperstimulation and multiple pregnancies. However, its efficacy remains low and the phenotype of women responding to this regimen needs to be characterized. Studies have shown that metformin
26
L. Duranteau et al. / Annales d’Endocrinologie 71 (2010) 25–27
Table 1 Metformin for regulating menstrual cycles. Recommendation no. 1. Rôle de la metformine pour régulariser les cycles. Recommandation no 1. Commentaries Metformin is not efficient enough in order to regulate menstrual cycles
Weak evidence in the literature (different criteria for the definition of PCOS, different doses, duration of treatment usually less than 6 months, no precise end-points, low number of patients included in each study) Efficient in 50% of treated patients, using the dose of 1500 mg/day No comparative study with weight loss Lower efficiency as compared to progestin treatment Responders to be determined
Table 4 Metformin in infertility. Recommendation no. 4. Rôle de la metformine en infertilité dans le syndrome des ovaires polykystiques. Recommandation no 4. Commentaries Metformin in addition to CC is not recommended as a second line treatment, after the failure of CC alone
CC: clomiphene citrate. Table 5 Metformin and gestational diabetes. Recommendation no. 5. Metformine et diabète gestationnel (DG). Recommandation no 5.
PCOS: polycystic ovary syndrome.
Commentaries
Table 2 Metformin in the treatment of hyperandrogenism. Recommendation no. 2. Rôle de la metformine pour traiter l’hyperandrogénie. Recommandation no 2.
Metformin should not be used during pregnancy in non diabetic women with PCOS, in order to prevent the risk of gestational diabetes (GD)
Commentaries Metformin is not efficient enough in order to treat hyperandrogenism
No data comparing metformin to progestin or contraceptive pills on large numbers of patients Less efficient than contraceptive pills and/or antiandrogens Potential low tolerance on the long term
Table 3 Metformin in infertility in polycystic ovary syndrome. Recommendation no. 3. Rôle de la metformine en infertilité dans le syndrome des ovaires polykystiques. Recommandation no 3. Commentaries Metformin should not be used as a first-line treatment in order to treat infertility Clomiphene citrate is the reference treatment
Studies are necessary in order to detect “good responders to metformin”: according to phenotype? Overweight Insulin resistance According to genotype?
Metformin increases the number of ovulation but does not increase the number of evolutive pregnancies 3 randomized studies on large cohorts of patients Weight loss remains useful
is not useful during in vitro fecundation (IVF) cycles as it does not increase the percentage of success. Its role in reducing pregnancy loss has not been confirmed in a recent meta-analysis [5] (Table 4). Another suggested use for metformin has been the prevention of gestational diabetes mellitus (GDM). Indeed, the frequency of GDM in women with PCOS is higher than in the general population (odds ratio: 2.89; IC: 1.68–4.98). Few randomized studies have tested the effect of metformin on the risk of GDM. The latest does not confirm a beneficial effect [6]. Furthermore, few data are available on the continuation of metformin during pregnancy, although no data are in favour of a harmful effect. It seems reasonable to stop metformin as soon as the diagnosis of pregnancy is performed (Table 5).
Randomized studies are needed in order to prove that metformin prevents GD No inferiority of metformin versus insulin when declared GD Safety of metformin during pregnancy to be proven
PCOS: polycystic ovary syndrome; GD: gestational diabetes.
On a long term basis, metformin could potentially reduce the risk of type 2 diabetes. The Diabetes Prevention Trial has shown a beneficial effect of metformin, reducing the risk of type 2 diabetes by 31%, as compared to placebo. However, few studies have specifically focused on the population of women with PCOS. As recommended by the American Diabetes Association (ADA), metformin should be prescribed in patients with high risk of diabetes and obesity, below the age of 60 [7]. As PCOS represents a risk factor for diabetes, metformin should be prescribed to obese patients with PCOS. However, metformin does not have significant effects on reducing weight [8]. In women with glucose intolerance or diabetes, it represents the best molecule, in addition to lifestyle modifications (Table 6). Metformin could potentially reduce the cardiovascular risk as women with PCOS present an increased risk [9]. However, new Table 6 Metformin in cardiovascular prevention. Recommendation no. 6. Rôle dans la prévention cardiovasculaire. Recommandation no 6. Commentaries Metformin should be prescribed to PCOS women when they are diabetic, in order to prevent their cardiovascular risk, after lifestyle modification
PCOS: polycystic ovary syndrome.
Insulin resistance in PCOS women is close to the insulin resistance observed in diabetic women without PCOS. Metformin has no residual effect after stopping the drug.
L. Duranteau et al. / Annales d’Endocrinologie 71 (2010) 25–27 Table 7 Metformin in cardiovascular prevention. Recommendation no. 7. Rôle dans la prévention cardiovasculaire. Recommandation no 7. Commentaries Metformin should not be used in PCOS non diabetic women in order to lose weight Metformin should not be used in order to treat dyslipidemia in women with PCOS
Metformin has no impact on weight nor on the waist/hip ratio, according to Lord’s meta-analysis Metformin has a low impact on LDL cholesterol.
PCOS: polycystic ovary syndrome; LDL: low density lipoproteins. Table 8 Metformin in cardiovascular prevention. Recommendation no. 8. Rôle dans la prévention cardiovasculaire. Recommandation no 8. Commentaries In PCOS women, without diabetes, but with fasting hyperglycemia or carbohydrate intolerance, metformin should be prescribed if BMI > 35 = ADA Recommendations
Metformin actions demonstrated on clinical and biological markers of cardiovascular risks Diabetes Prevention Program (DPP) has shown a reduction in the risk of developing a diabetes if intensive coaching, diet and/or metformin with a beneficial cost effect But: No study of mortality in PCOS Data are extrapolated from UKPDS No authority approval in France for using metformin apart from diabetic women.
PCOS: polycystic ovary syndrome; BMI: body mass index; UKPDS: United Kingdom Prospective Diabetes Study; ADA: American Diabetes Assocation; DPP: Diabetes Prevention Program.
data have shown that this cardiovascular risk differs according to the phenotype of the syndrome [10]. So far, no proof has been given as to a beneficial effect of metformin in reducing the cardiovascular risk, in women with PCOS. However, some data can be extrapolated from the UKPDS (United Kingdom Prospective Diabetes Study) trial. This study has shown a continued benefit of metformin on the cardiovascular risk in overweight diabetic patients, during ten years of post-trial follow-up [11] (Tables 7 and 8).
27
In summary, metformin is not the best treatment in order to regularise menstrual cycle, nor to treat hyperandrogenism. It is not as effective as CC as first-line therapy for women with PCOS, for ovulation induction. It does not reduce the risk of pregnancy loss. However, it may be beneficial in order to reduce the risk of diabetes in overweight women with PCOS and to reduce the cardiovascular risk in those women with glucose intolerance or type 2 diabetes. 1. French version A French version of this article is available at doi: 10.1016/j.ando.2009.12.006 References [1] Rotterdam ESHRE/ASRM sponsored PCOS consensus workshop group. Hum Reprod 2004;19:41–7. [2] Costello M, Shrestha B, Eden J, Sjoblom P, Johnson N. Insulinsensitizing drugs versus the combined oral contraceptive pill for hirsutism, acne and risk of diabetes, cardiovascular disease, and endometrial cancer in polycystic ovary syndrome. Cochrane Database Syst Rev 2007: CD005552. [3] Palomba S, Falbo A, Zullo F, Orio F. Evidence-based and potential benefits of metformin in polycystic ovary syndrome. Endocr Rev 2009;30: 1–50. [4] Creanga AA, Bradley HM, McCormick C, Witkop CT. Uses of metformin in polycystic ovary syndrome: a meta analysis. Obstet Gynecol 2008;111:959–68. [5] Palomba S, Falbo A, Orio F, Zullo F. Preconceptional effects of metformin on abortion risk in polycystic ovary syndrome (PCOS): a systematic review and meta-analysis of randomized controlled trials (RCTs). Fertil Steril 2008 epub ahead of print. [6] Fougner KJ, Vanky E, Carlsen SM. Metformin has no major effects on glucose homeostasis in pregnant women with PCOS: results of a randomized double-blind study. Scand J Clin Lab Invest 2008;24:1–6. [7] American Diabetes Association Standards for medical care in diabetes2009 Diabetes Care 2009;32:13–61. [8] Lord JM, Flight IH, Norman RJ. Insulin-sensitizing drugs (metformin, troglitazone, rosiglitazone, pioglitazone, D-chiro-inosital) for polycystic ovary syndrome. Cochrane Database Syst Rev 2003:CD003053. [9] Rizzo M, Berneis K, Spinas G, Rini GB, Carmina E. Long-term consequence of polycystic ovary syndrome on cardiovascular risk. Fertil Steril 2009;91:1563–7. [10] Jovanovic VP, Carmina E, Lobo RA. Not all women diagnosed with PCOS share the same cardiovascular risk profile Fertil Steril 2009 June 4 epub ahead of print. [11] Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA. 10-year follow-up of intensive glucose control in type 2 diabetes. N Engl J Med 2008;359:1577–89.
Consensus of French Endocrine Society on female hyperandrogenism
Should physicians prescribe metformin to women with polycystic ovary syndrome PCOS? Rôle de la metformine dans le SOPK L. Duranteau a,∗ , P. Lefevre b , N. Jeandidier c , T. Simon d , S. Christin-Maitre e a
Service d’endocrinologie pédiatrique, hôpital Saint-Vincent-de-Paul, 82, avenue Denfert-Rochereau, 75014 Paris, France b Service d’endocrinologie, hôpital Lapeyronie, 34925 Montpellier, France c Service d’endocrinologie, Hospices civils, 67091 Strasbourg, France d Service de pharmacologie, Urcest, hôpital Saint-Antoine, 75571 Paris, France e Service d’endocrinologie, hôpital Saint-Antoine, Paris, France Available online 15 January 2010
Abstract 1. Metformin is not efficient enough in order to regulate menstrual cycles. 2.Metformin is not efficient enough in order to treat hyperandrogenism. 3. Metformin should not be used as a first-line treatment in order to treat infertility. Clomiphene citrate (CC) is the reference treatment. 4. Metformin in addition to CC is not recommended as a second line treatment, after the failure of CC alone. 5. Metformin should not be used during pregnancy in non diabetic women with PCOS, in order to prevent the risk of gestational diabetes. 6. Metformin should be prescribed to PCOS women when they are diabetic, in order to prevent their cardiovascular risk, after lifestyle modification. 7. Metformin should not be used in PCOS non diabetic women in order to lose weight. Metformin should not be used in order to treat dyslipidemia in women with PCOS. 8. In PCOS women, without diabetes, but with fasting hyperglycemia or carbohydrate intolerance, metformin should be prescribed if: BMI > 35. © 2009 Published by Elsevier Masson SAS. Keywords: Metformin; Polycystic ovary syndrome; French consensus
Metformin has been used in Europe since the seventies in order to treat type 2 diabetes. It has been prescribed in the United States since 1994. This molecule improves insulin resistance as it increases glucose capture in muscles and decreases neoglucogenesis in the liver. Women with polycystic ovary syndrome (PCOS) present insulin resistance with hyperinsulinemia in 50 to 70% of cases [1]. Therefore, in theory, metformin represents a good candidate to treat PCOS. The use of metformin has been studied in order to regulate cycle disturbances. It regulates menstrual frequency, using doses ranging from 1000 to 2500 mg per day. However, only few randomized studies have been performed and the duration of treatment reached a maximum of one year. Metformin’s efficacy remains lower than combined oral contraceptive pill [2].
∗
DOI of original article:10.1016/j.ando.2009.12.006. Corresponding author. E-mail address: [email protected] (L. Duranteau).
0003-4266/$ – see front matter © 2009 Published by Elsevier Masson SAS. doi:10.1016/j.ando.2009.12.005
Progestins alone remain the treatment of reference in order to regulate menstrual bleeding (Table 1). Metformin has been used in order to reduce hyperandrogenism and hirsutism. According to different meta-analysis, its effect is not higher than placebo and lower than antiandrogen treatments [3] (Table 2). In order to treat infertility, many hopes arrived in the late nineties, concerning metformin treatment. However, most recent randomized studies performed on large groups of patients have demonstrated that it induces a lower number of live births than clomiphene citrate (CC) [4]. Therefore, CC should remain the first line treatment in order to induce ovulation and pregnancy in women with PCOS, after lifestyle modifications (Table 3). In CC resistant women, metformin has been added to CC. This treatment has a low cost and induces few ovarian hyperstimulation and multiple pregnancies. However, its efficacy remains low and the phenotype of women responding to this regimen needs to be characterized. Studies have shown that metformin
26
L. Duranteau et al. / Annales d’Endocrinologie 71 (2010) 25–27
Table 1 Metformin for regulating menstrual cycles. Recommendation no. 1. Rôle de la metformine pour régulariser les cycles. Recommandation no 1. Commentaries Metformin is not efficient enough in order to regulate menstrual cycles
Weak evidence in the literature (different criteria for the definition of PCOS, different doses, duration of treatment usually less than 6 months, no precise end-points, low number of patients included in each study) Efficient in 50% of treated patients, using the dose of 1500 mg/day No comparative study with weight loss Lower efficiency as compared to progestin treatment Responders to be determined
Table 4 Metformin in infertility. Recommendation no. 4. Rôle de la metformine en infertilité dans le syndrome des ovaires polykystiques. Recommandation no 4. Commentaries Metformin in addition to CC is not recommended as a second line treatment, after the failure of CC alone
CC: clomiphene citrate. Table 5 Metformin and gestational diabetes. Recommendation no. 5. Metformine et diabète gestationnel (DG). Recommandation no 5.
PCOS: polycystic ovary syndrome.
Commentaries
Table 2 Metformin in the treatment of hyperandrogenism. Recommendation no. 2. Rôle de la metformine pour traiter l’hyperandrogénie. Recommandation no 2.
Metformin should not be used during pregnancy in non diabetic women with PCOS, in order to prevent the risk of gestational diabetes (GD)
Commentaries Metformin is not efficient enough in order to treat hyperandrogenism
No data comparing metformin to progestin or contraceptive pills on large numbers of patients Less efficient than contraceptive pills and/or antiandrogens Potential low tolerance on the long term
Table 3 Metformin in infertility in polycystic ovary syndrome. Recommendation no. 3. Rôle de la metformine en infertilité dans le syndrome des ovaires polykystiques. Recommandation no 3. Commentaries Metformin should not be used as a first-line treatment in order to treat infertility Clomiphene citrate is the reference treatment
Studies are necessary in order to detect “good responders to metformin”: according to phenotype? Overweight Insulin resistance According to genotype?
Metformin increases the number of ovulation but does not increase the number of evolutive pregnancies 3 randomized studies on large cohorts of patients Weight loss remains useful
is not useful during in vitro fecundation (IVF) cycles as it does not increase the percentage of success. Its role in reducing pregnancy loss has not been confirmed in a recent meta-analysis [5] (Table 4). Another suggested use for metformin has been the prevention of gestational diabetes mellitus (GDM). Indeed, the frequency of GDM in women with PCOS is higher than in the general population (odds ratio: 2.89; IC: 1.68–4.98). Few randomized studies have tested the effect of metformin on the risk of GDM. The latest does not confirm a beneficial effect [6]. Furthermore, few data are available on the continuation of metformin during pregnancy, although no data are in favour of a harmful effect. It seems reasonable to stop metformin as soon as the diagnosis of pregnancy is performed (Table 5).
Randomized studies are needed in order to prove that metformin prevents GD No inferiority of metformin versus insulin when declared GD Safety of metformin during pregnancy to be proven
PCOS: polycystic ovary syndrome; GD: gestational diabetes.
On a long term basis, metformin could potentially reduce the risk of type 2 diabetes. The Diabetes Prevention Trial has shown a beneficial effect of metformin, reducing the risk of type 2 diabetes by 31%, as compared to placebo. However, few studies have specifically focused on the population of women with PCOS. As recommended by the American Diabetes Association (ADA), metformin should be prescribed in patients with high risk of diabetes and obesity, below the age of 60 [7]. As PCOS represents a risk factor for diabetes, metformin should be prescribed to obese patients with PCOS. However, metformin does not have significant effects on reducing weight [8]. In women with glucose intolerance or diabetes, it represents the best molecule, in addition to lifestyle modifications (Table 6). Metformin could potentially reduce the cardiovascular risk as women with PCOS present an increased risk [9]. However, new Table 6 Metformin in cardiovascular prevention. Recommendation no. 6. Rôle dans la prévention cardiovasculaire. Recommandation no 6. Commentaries Metformin should be prescribed to PCOS women when they are diabetic, in order to prevent their cardiovascular risk, after lifestyle modification
PCOS: polycystic ovary syndrome.
Insulin resistance in PCOS women is close to the insulin resistance observed in diabetic women without PCOS. Metformin has no residual effect after stopping the drug.
L. Duranteau et al. / Annales d’Endocrinologie 71 (2010) 25–27 Table 7 Metformin in cardiovascular prevention. Recommendation no. 7. Rôle dans la prévention cardiovasculaire. Recommandation no 7. Commentaries Metformin should not be used in PCOS non diabetic women in order to lose weight Metformin should not be used in order to treat dyslipidemia in women with PCOS
Metformin has no impact on weight nor on the waist/hip ratio, according to Lord’s meta-analysis Metformin has a low impact on LDL cholesterol.
PCOS: polycystic ovary syndrome; LDL: low density lipoproteins. Table 8 Metformin in cardiovascular prevention. Recommendation no. 8. Rôle dans la prévention cardiovasculaire. Recommandation no 8. Commentaries In PCOS women, without diabetes, but with fasting hyperglycemia or carbohydrate intolerance, metformin should be prescribed if BMI > 35 = ADA Recommendations
Metformin actions demonstrated on clinical and biological markers of cardiovascular risks Diabetes Prevention Program (DPP) has shown a reduction in the risk of developing a diabetes if intensive coaching, diet and/or metformin with a beneficial cost effect But: No study of mortality in PCOS Data are extrapolated from UKPDS No authority approval in France for using metformin apart from diabetic women.
PCOS: polycystic ovary syndrome; BMI: body mass index; UKPDS: United Kingdom Prospective Diabetes Study; ADA: American Diabetes Assocation; DPP: Diabetes Prevention Program.
data have shown that this cardiovascular risk differs according to the phenotype of the syndrome [10]. So far, no proof has been given as to a beneficial effect of metformin in reducing the cardiovascular risk, in women with PCOS. However, some data can be extrapolated from the UKPDS (United Kingdom Prospective Diabetes Study) trial. This study has shown a continued benefit of metformin on the cardiovascular risk in overweight diabetic patients, during ten years of post-trial follow-up [11] (Tables 7 and 8).
27
In summary, metformin is not the best treatment in order to regularise menstrual cycle, nor to treat hyperandrogenism. It is not as effective as CC as first-line therapy for women with PCOS, for ovulation induction. It does not reduce the risk of pregnancy loss. However, it may be beneficial in order to reduce the risk of diabetes in overweight women with PCOS and to reduce the cardiovascular risk in those women with glucose intolerance or type 2 diabetes. 1. French version A French version of this article is available at doi: 10.1016/j.ando.2009.12.006 References [1] Rotterdam ESHRE/ASRM sponsored PCOS consensus workshop group. Hum Reprod 2004;19:41–7. [2] Costello M, Shrestha B, Eden J, Sjoblom P, Johnson N. Insulinsensitizing drugs versus the combined oral contraceptive pill for hirsutism, acne and risk of diabetes, cardiovascular disease, and endometrial cancer in polycystic ovary syndrome. Cochrane Database Syst Rev 2007: CD005552. [3] Palomba S, Falbo A, Zullo F, Orio F. Evidence-based and potential benefits of metformin in polycystic ovary syndrome. Endocr Rev 2009;30: 1–50. [4] Creanga AA, Bradley HM, McCormick C, Witkop CT. Uses of metformin in polycystic ovary syndrome: a meta analysis. Obstet Gynecol 2008;111:959–68. [5] Palomba S, Falbo A, Orio F, Zullo F. Preconceptional effects of metformin on abortion risk in polycystic ovary syndrome (PCOS): a systematic review and meta-analysis of randomized controlled trials (RCTs). Fertil Steril 2008 epub ahead of print. [6] Fougner KJ, Vanky E, Carlsen SM. Metformin has no major effects on glucose homeostasis in pregnant women with PCOS: results of a randomized double-blind study. Scand J Clin Lab Invest 2008;24:1–6. [7] American Diabetes Association Standards for medical care in diabetes2009 Diabetes Care 2009;32:13–61. [8] Lord JM, Flight IH, Norman RJ. Insulin-sensitizing drugs (metformin, troglitazone, rosiglitazone, pioglitazone, D-chiro-inosital) for polycystic ovary syndrome. Cochrane Database Syst Rev 2003:CD003053. [9] Rizzo M, Berneis K, Spinas G, Rini GB, Carmina E. Long-term consequence of polycystic ovary syndrome on cardiovascular risk. Fertil Steril 2009;91:1563–7. [10] Jovanovic VP, Carmina E, Lobo RA. Not all women diagnosed with PCOS share the same cardiovascular risk profile Fertil Steril 2009 June 4 epub ahead of print. [11] Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA. 10-year follow-up of intensive glucose control in type 2 diabetes. N Engl J Med 2008;359:1577–89.