Metformin for treatment of the polycystic ovary syndrome

INTRODUCTION — The polycystic ovary syndrome (PCOS) is characterized by both oligo/amenorrhea and androgen excess in women. When fully expressed, the manifestations include irregular menstrual cycles, hirsutism, obesity, and a constellation of cardiometabolic disturbances. It is a common endocrinopathy, occurring in 5 to 7 percent of reproductive age women [1-3].

The use of metformin in the management of PCOS will be reviewed here. The clinical manifestations, diagnosis, and other treatment options for PCOS are reviewed separately. (See "Clinical manifestations of polycystic ovary syndrome in adults" and "Diagnosis of polycystic ovary syndrome in adults" and "Treatment of polycystic ovary syndrome in adults".)

OVERVIEW — Interest in the use of metformin, an insulin-lowering drug, in PCOS increased when it was appreciated that insulin resistance played an important role in the pathophysiology of the disorder. However, we no longer recommend the use of metformin as a first-line therapy for any indication in women with PCOS. Metformin is typically the first-line treatment for patients with type 2 diabetes; it is not approved for use in prediabetes or PCOS, although it is often prescribed for treatment of these conditions.

Early trials in women with PCOS subsequently demonstrated a small benefit for weight reduction, a decrease in serum androgens (without improvement in hirsutism), and restoration of menstrual cycles in approximately 50 percent of women with oligomenorrhea (although not always ovulatory). Early data also suggested that metformin was effective for ovulation induction in anovulatory women with PCOS. As a result, metformin was used "off-label" for a number of these indications [4,5].

However, available data do not support the use of metformin for treatment of hirsutism or as first-line treatment for ovulation induction in this population. However, there may be clinical situations where metformin might be used as an alternative to other first-line treatments for hirsutism (combined estrogen-progestin oral contraceptives [COCs] and antiandrogens or ovulation induction [letrozole or clomiphene]) when there is a contraindication to those agents or the patient prefers not to use a first-line agent.

Mechanisms of action — Metformin's major effect is to decrease hepatic glucose production, thus reducing the need for insulin secretion; it also decreases intestinal absorption of glucose. Metformin also has an antilipolytic effect that lowers free fatty acid concentrations, thus reducing gluconeogenesis [6,7].

Pharmacology — Metformin is a biguanide antihyperglycemic approved for management of type 2 diabetes mellitus when hyperglycemia cannot be managed by diet and exercise alone. Unlike sulfonylureas, it does not produce hypoglycemia in either normal subjects or patients with type 2 diabetes. Metformin is also suggested for diabetes prevention in patients age <60 years and/or body mass index (BMI) ≥35 kg/m², women with a history of gestational diabetes mellitus (GDM) with impaired fasting glucose (IFG), impaired glucose tolerance (IGT), or glycated hemoglobin (A1C) of 5.7 to 6.4 percent, in whom lifestyle interventions failed to improve glycemic indices. (See "Metformin in the treatment of adults with type 2 diabetes mellitus" and "Prevention of type 2 diabetes mellitus", section on 'Metformin'.)

Metformin is rapidly absorbed from the small intestine with peak plasma levels occurring two hours after ingestion. Food decreases the rate of drug absorption and peak drug concentration, although taking the medication with food is recommended to attenuate its gastrointestinal side effects. Metformin is not metabolized and is largely excreted in the urine. Renal clearance is approximately 3.5 times greater than renal clearance of creatinine, which indicates that tubular secretion is the main route of excretion. The plasma elimination half-life is approximately six hours.

Dose — Metformin is available in a generic form as 500, 850, and 1000 mg tablets. The target dose is 1500 to 2000 mg daily; clinically significant responses are not regularly observed at doses less than 1000 mg daily. Many clinicians begin treatment with 500 mg taken with a meal to reduce gastrointestinal side effects. If tolerated, the dose can be increased to 500 mg at both lunch and dinner and then to 500 mg at breakfast, lunch, and dinner. One to two weeks should elapse between increases in dose.

Extended-release tablets are available in a generic form as 500, 750, and 1000 mg tablets. Extended-release metformin may be associated with fewer side effects, and in general, we suggest starting with these formulations rather than the short-acting ones. With extended-release tablets, the entire daily dose may be given at dinner time. The dose is initially 500 mg with dinner and is escalated gradually to a maximum of 2000 to 2250 mg.

Finally, metformin exists in both a liquid formulation (500 mg/5 mL) and a specially formulated tablet designed to release the medication more slowly into the intestinal tract [8]. In instances where the traditional formulations are not tolerated, it is occasionally helpful to have the patient try either the liquid or slow-release formulation before abandoning metformin therapy.

Side effects and precautions — The most common side effects are gastrointestinal: diarrhea, nausea or vomiting, flatulence, indigestion, and abdominal discomfort. These symptoms are usually mild, transient, and reversible after dose reduction or discontinuation of the drug.

Metformin also reduces intestinal absorption of vitamin B12 in up to 30 percent of patients and lowers serum vitamin B12 concentrations in 5 to 10 percent but only rarely causes megaloblastic anemia [9]. In women with PCOS taking metformin, we suggest yearly monitoring for B12 deficiency with a complete blood count (CBC) and serum B12 concentration. In addition, women should make sure they meet the recommended dietary allowances for B12 (2.6 mcg for adults). This can be done either with food or supplements. (See "Metformin in the treatment of adults with type 2 diabetes mellitus", section on 'Vitamin B12 deficiency' and "Clinical manifestations and diagnosis of vitamin B12 and folate deficiency".)

Lactic acidosis has been described, but it is an extremely rare complication in otherwise healthy individuals. However, metformin should not be prescribed for women with other conditions that increase the risk of lactic acidosis, such as renal insufficiency, congestive heart failure, or sepsis. (See "Metformin in the treatment of adults with type 2 diabetes mellitus", section on 'Lactic acidosis'.)

Breastfeeding — Although data are limited, metformin use by breastfeeding women with PCOS does not appear to be associated with adverse outcomes. Concentrations of metformin in breast milk appear to be low, and the mean infant exposure to metformin has been reported to be well below the level of concern for breastfeeding; no adverse effects on blood glucose of nursing infants have been reported [10].

POTENTIAL USES — As discussed above, treatment of type 2 diabetes mellitus is the only approved indication for metformin. Nevertheless, it has been used "off-label" to treat or prevent several clinical problems associated with PCOS, including oligomenorrhea, hirsutism, anovulatory infertility, prevention of pregnancy complications, and obesity [4,11]. However, it is not a first-line drug for oligomenorrhea, and it is ineffective for most other indications. (See 'No longer indicated' below.)

Although it might be preferable to treat only insulin-resistant women with PCOS, there is currently no readily available, reliable test to evaluate the presence of insulin resistance. Some clinicians do not test at all; others perform an oral glucose tolerance test to look for impaired glucose tolerance (IGT) and check for elevated insulin levels. We do not suggest measuring insulin levels, as the results do not affect management. (See "Insulin resistance: Definition and clinical spectrum" and "Diagnosis of polycystic ovary syndrome in adults", section on 'Cardiometabolic risk assessment'.)

Reproductive

Oligomenorrhea — For women with oligomenorrhea due to PCOS who need endometrial protection (and contraception or management of hirsutism), we suggest combined estrogen-progestin oral contraceptives (COCs) as first-line therapy. Therapeutic effects include the following (see "Treatment of polycystic ovary syndrome in adults", section on 'Menstrual dysfunction'):

●A decrease in luteinizing hormone (LH) secretion and a resulting decrease in ovarian androgen production

●An increase in hepatic production of sex hormone-binding globulin (SHBG) with a decrease in bioavailable testosterone

●A decrease in adrenal androgen secretion

●Regular withdrawal bleeding

●Prevention of endometrial hyperplasia

For second-line therapy and for situations where COCs are contraindicated, we suggest continuous or cyclic progestin-only treatment, or a levonorgestrel-releasing intrauterine device (IUD) for endometrial protection. (See "Contraception: Progestin-only pills (POPs)" and "Intrauterine contraception: Background and device types", section on 'Levonorgestrel IUD'.)

Metformin may also be used as second-line therapy (particularly in women with contraindications to COC use), but it has not been proven to be endometrial protective.

Metformin will restore ovulatory menses in approximately 50 percent of women with PCOS [12-14], although some studies report ovulatory rates from 23 to 90 percent [15-21]. Evidence for the efficacy of metformin includes the following:

●In one trial, 23 women with PCOS were randomly assigned to receive metformin (500 mg three times per day) or placebo for six months [12]. Approximately one-half of the women treated with metformin achieved normalization of menstrual function, confirmed by intermenstrual interval and luteal phase serum progesterone monitoring.

●In a meta-analysis of 13 trials, women treated with metformin had a fourfold higher chance of ovulating than women treated with placebo [22]. A similar fourfold increase was seen for the combination of metformin plus clomiphene when compared with clomiphene alone. The use of metformin for ovulation induction is reviewed below. (See 'Anovulatory infertility' below.)

●Endometrial protection – Some women do not ovulate during the first six months of treatment, and therefore, we recommend cyclic progestin treatment during this interval. Although some women may have restoration of menses with metformin, unless cycles are regular and ovulatory, one cannot assume that the patient has reliable endometrial protection. For those who have menses that are irregular and anovulatory, cyclic progestins or COCs are necessary. We typically use either micronized progesterone (200 mg/day for 14 days every one to two months) or medroxyprogesterone acetate (10 mg/day for 12 to 14 days every one to two months).

●Women who do not attain regular ovulatory menses should be treated with cyclic progestin therapy to reduce the risk of endometrial hyperplasia until regular menses are established. (See "Endometrial hyperplasia: Clinical features, diagnosis, and differential diagnosis".)

During the initial phase of oligomenorrhea, it may be difficult to determine if ovulation has taken place. Measuring serum progesterone every 10 days may be one approach, although cumbersome, to identify if ovulation has occurred following initiation of metformin therapy.

We do not typically use this approach. If menstrual cycles are regular, we do not check serum progesterone levels. If cycles are irregular, we add cyclic progestin therapy as outlined above. If the patient then develops regular menses between courses of cyclic progestin therapy, the progestin can be stopped to see if spontaneous ovulations have developed.

A contraceptive (eg, diaphragm, condom) should be prescribed for all sexually active women not planning pregnancy since metformin treatment may induce ovulation.

●COCs versus metformin – There have been concerns that estrogen-progestin contraceptives may worsen insulin resistance in women with PCOS and that metformin should therefore be the preferred treatment. However, while COCs are less beneficial than metformin for insulin sensitivity, they are more effective for androgen suppression and menstrual cycle control. In addition, available data have not shown any evidence of adverse metabolic risk [23].

In a meta-analysis of four trials that included 104 women with PCOS, metformin was less effective than COCs for improving menstrual pattern and in reducing serum total testosterone concentration but more effective for reducing fasting insulin and not increasing fasting triglyceride concentrations [24]. (See "Treatment of polycystic ovary syndrome in adults", section on 'Metabolic effects of COCs in PCOS'.)

●Dosing – The optimum dose of metformin to restore ovulatory menses has not been determined. The two best studied metformin regimens are 500 mg three times daily (1500 mg daily) and 850 mg twice daily (1700 mg daily). Some clinicians go up to 2000 mg daily [25]. Extended-release preparations can also be used (500 mg tablets, with the entire dose of 1500 to 2000 mg administered with dinner).

IVF pretreatment — In women with PCOS, the administration of metformin before or during in vitro fertilization (IVF) cycles does not appear to improve clinical pregnancy or live-birth rates [26,27]. However, metformin administration before or during IVF cycles does appear to reduce the number of retrieved oocytes [28] and the risk of ovarian hyperstimulation syndrome (OHSS) [26,27,29]. Similar estimates of benefits were reported in two meta-analyses (odds ratio [OR] 0.27 for risk of OHSS with metformin compared with placebo use). Although metformin therapy is a potential option to lower the risk of OHSS with IVF, there appear to be more effective strategies, such as gonadotropin-releasing hormone (GnRH) antagonists [30]. The addition of metformin in this setting does not appear to lower the risk of multiple gestation [31]. (See "Prevention of ovarian hyperstimulation syndrome".)

Metabolic — In a trial described above [12], compared with placebo, the metformin group had improvement in plasma insulin and insulin sensitivity (as measured by glucose clamp studies), a reduction in serum free testosterone, and an increase in mean serum high-density lipoprotein (HDL) cholesterol. Both the biochemical and clinical changes were independent of changes in body weight and were sustained in a follow-up, open, long-term, observational study (mean duration of treatment was 11 months). However, there was no significant reduction in hirsutism, despite the lower free serum testosterone concentration. (See 'Hirsutism' below.)

A second 14-week trial in 94 women with PCOS randomly assigned to metformin or placebo showed similar clinical and metabolic benefits, except in the most obese women (body mass index [BMI] >37 kg/m²) [15]. A meta-analysis of 13 trials reported that metformin is associated with a reduction in blood pressure, low-density lipoprotein (LDL), and fasting insulin levels, when compared with placebo [32].

Prevention of diabetes mellitus — Women with PCOS are at increased risk for developing type 2 diabetes mellitus or IGT. In a study of 122 obese women with PCOS, 45 percent had either IGT (35 percent) or type 2 diabetes mellitus (10 percent) by age 40 [33]. As is the case in non-PCOS populations, the risk of type 2 diabetes in PCOS appears to be highest among those who are obese and/or have a significant family history of type 2 diabetes [33]. The frequency of diagnosis of type 2 diabetes in women with PCOS varies with the screening test used. A fasting blood glucose may underestimate the prevalence of diabetes when compared with an oral glucose tolerance test. (See "Clinical manifestations of polycystic ovary syndrome in adults", section on 'IGT/type 2 diabetes'.)

Retrospective data in women with PCOS suggest that metformin therapy may delay or prevent conversion to IGT and diabetes [34]. However, in the absence of clinical trial data, we do not recommend metformin for women with normal glucose tolerance for this indication.

NO LONGER INDICATED

Hirsutism — The first-line therapy for treatment of hirsutism caused by PCOS is a combined estrogen-progestin oral contraceptive (COC), as recommended by the 2018 Endocrine Society Guidelines. An antiandrogen is then added after six months if the cosmetic response is suboptimal. (See "Management of hirsutism in premenopausal women".)

A number of clinical trials have examined the effects of insulin-lowering agents on circulating androgen concentrations in women with PCOS and have noted a reduction in these hormones with therapy [12,35-41]. Although metformin may reduce serum androgen concentrations, it has limited benefit for the treatment of hirsutism [23,42-48].

Anovulatory infertility — A structured approach, starting with low-cost interventions and advancing to high-resource interventions, is warranted for treatment of anovulatory infertility in women with PCOS (table 1). The first two management steps are weight loss through caloric restriction and increased exercise for women with a body mass index (BMI) >27 kg/m², followed by medical therapy. (See "Treatment of polycystic ovary syndrome in adults", section on 'Weight loss'.)

First-line options for ovulation induction in anovulatory women with PCOS include letrozole and clomiphene. The best available evidence suggests that letrozole and clomiphene are more effective than metformin for live-birth rates, the most important outcome. Early ovulation induction trials suggested better results with metformin alone or metformin plus clomiphene when compared with clomiphene alone [12,13,15-19,32,36,49-53].

However, data from a number of multicenter trials have now reported that while metformin may be effective for restoring ovulation, it appears to be less effective for fertility (live-birth rates) when compared with clomiphene [54-60].

Letrozole, an aromatase inhibitor, is also effective for ovulation induction in women with PCOS (although it is not approved for this indication). Available data suggest that live-birth rates are higher with letrozole than clomiphene, and many experts now suggest letrozole as first-line therapy for anovulatory women with PCOS. (See "Ovulation induction with letrozole".)

In a 2012 meta-analysis of 38 metformin trials in 3495 women, there was no evidence that metformin improved live-birth rates, whether it was used alone (pooled odds ratio [OR] 1.80, 95% CI 0.52-6.16, three trials, 115 women) or in combination with clomiphene (pooled OR 1.16, 95% CI 0.85-1.56, seven trials, 907 women) [22]. Thus, larger trials and a meta-analysis suggest that clomiphene is more effective than metformin for live-birth rates, the most relevant outcome [54], and that the addition of metformin to clomiphene does not provide additional benefit [54,55]. However, in one of the trials, multiple pregnancy rates were higher with clomiphene [54].

A consensus group has recommended against the routine use of metformin for ovulation induction except in women with glucose intolerance [61]. In addition, the American Society for Reproductive Medicine (ASRM) has published a summary of the available evidence on the role of metformin for ovulation induction with PCOS [31]. Although metformin was widely used in the past for women with PCOS for ovulation induction and prevention of miscarriage, multiple gestations, and pregnancy complications, the ASRM concludes that there is insufficient evidence to recommend metformin for any of these indications. Metformin does increase ovulatory rates and pregnancy rates, but unlike clomiphene and letrozole, it does not improve live birth rates and is therefore not the drug of choice. (See "Treatment of polycystic ovary syndrome in adults", section on 'Ovulation induction medications'.)

Prevention of pregnancy complications

Spontaneous abortion — Small observational studies had shown benefits of metformin therapy for reducing the risk of early miscarriage in pregnant women with PCOS, but randomized trials have shown no benefit.

The spontaneous abortion rate in women with PCOS is 20 to 40 percent higher than the baseline in the general obstetric population [62]. Three early studies in women with PCOS reported miscarriage rates of 62 to 73 percent in pregnancies when metformin was not taken and 9 to 36 percent of pregnancies in the same women when metformin was taken [62-65].

However, in a meta-analysis of 17 trials in women with PCOS taking metformin with or without clomiphene for ovulation induction, no effect of metformin was seen on the risk of pregnancy loss [66].

Gestational diabetes — Early studies also suggested that metformin therapy reduced the risk of gestational diabetes mellitus (GDM), but this was not confirmed in randomized trials.

The incidence of GDM is severalfold higher in women with PCOS than in the general obstetric population [65,67]. Two observational studies suggested that metformin use was associated with a lower risk of GDM, but this was not confirmed in a trial of 273 pregnancies among 257 women with PCOS who were randomly assigned to receive metformin (2000 mg/day) or placebo from the first trimester until delivery. There was no significant difference in prevalence of GDM between groups (metformin 17.6 percent versus placebo 16.9 percent) [68].

The treatment of established GDM is discussed separately. (See "Gestational diabetes mellitus: Glucose management and maternal prognosis", section on 'Metformin'.)

Other complications — In the same randomized trial of metformin versus placebo during pregnancy, no significant differences were seen in the prevalence of preeclampsia (7.4 versus 3.7 percent) or preterm delivery (3.7 versus 8.2 percent) or a composite outcome that included preeclampsia, preterm delivery, and GDM (overall rate of complications 25.9 and 24.4 percent for metformin and placebo, respectively) [68]. Women taking metformin gained less weight than women in the placebo group (mean difference -2.2 kg).

Neonatal outcomes — A meta-analysis of pregnancy outcome in 172 women after first-trimester exposure to metformin did not find an increased risk of major malformations when compared with control women (OR 0.50, 95% CI 0.15-1.60) [69]. (See "Gestational diabetes mellitus: Glucose management and maternal prognosis", section on 'Metformin'.)

However, in utero exposure to metformin from the first trimester to birth among women with PCOS may be associated with excess weight of those children at 4 years of age [70].

Obesity — At least 50 percent of women with PCOS are obese. For obese women with PCOS who need to lose weight, we suggest other weight loss medications or bariatric surgery rather than metformin. Some benefit is seen when metformin is combined with caloric restriction [71-78], but metformin does not produce enough weight loss to qualify as a weight loss drug. We agree with the Endocrine Society 2013 guidelines and do not suggest metformin for women with PCOS and obesity unless they have impaired glucose tolerance and have failed lifestyle interventions. (See "Obesity in adults: Drug therapy" and 'Prevention of diabetes mellitus' above.)

ADDING METFORMIN TO ORAL CONTRACEPTIVES — There is currently inadequate evidence to recommend the routine addition of metformin to combined estrogen-progestin oral contraceptive (COC) therapy as it is still unclear whether this combination has important cosmetic or metabolic advantages over COC monotherapy.

●In one randomized clinical trial, 40 nonobese women with PCOS were treated with either a COC alone (ethinyl estradiol 35 mcg/day plus cyproterone acetate 2 mg/day) or this COC plus metformin (500 mg three times daily) for four months [79]. Combination therapy of the COC plus metformin resulted in the greatest reduction of the serum androstenedione level and the greatest increase in sex hormone-binding globulin (SHBG) (figure 1).

Weight loss occurred only in the group taking metformin. Circulating androgen concentrations and Ferriman-Gallwey hirsutism scores decreased to a similar degree in both groups. Thus, the combination of metformin plus a COC may be useful when weight loss is desired but not when used for other indications.

●In a second trial, 30 women with PCOS were randomly assigned to receive a COC alone (ethinyl estradiol 35 mcg/day plus norgestimate 25 mcg/day) or in combination with metformin (1500 mg/day) for six months [80].

Although obese women were not specifically excluded, the population was relatively lean (mean body mass index [BMI]±standard deviation [SD] 22.1±3.1 and 24.7±4.9 kg/m² in the monotherapy and combined therapy groups, respectively). At six months, the only significant difference between the groups was a lower free androgen index in the combination compared with the monotherapy group. No differences were seen in lipids, insulin sensitivity (measured by euglycemic hyperinsulinemic clamp technique), SHBG, or testosterone. It is possible that results would be different in an obese population.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Polycystic ovary syndrome" and "Society guideline links: Hirsutism".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Polycystic ovary syndrome (The Basics)")

●Beyond the Basics topics (see "Patient education: Polycystic ovary syndrome (PCOS) (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Use of metformin in PCOS – We no longer recommend the use of metformin as a first-line therapy for any indication in women with polycystic ovary syndrome (PCOS). However, it is still sometimes prescribed, and its use depends upon the individual patient's goals.

●Oligomenorrhea/endometrial protection – For women with oligomenorrhea who need endometrial protection, we suggest combined estrogen-progestin oral contraceptives (COCs) as first-line therapy rather than metformin. For second-line therapy and for situations where COCs are contraindicated, we suggest continuous or cyclic progestin-only treatment, or a levonorgestrel-secreting intrauterine device (IUD) for endometrial protection. Metformin may also be used as second-line therapy (particularly in women with contraindications to pill use), but it has not been proven to be endometrial protective. When metformin is used, cyclic progestin therapy may be added for the first six months of metformin treatment until regular cycles are established. (See 'Oligomenorrhea' above and "Treatment of polycystic ovary syndrome in adults", section on 'Women pursuing pregnancy'.)

●Hyperandrogenism – For management of hyperandrogenic symptoms alone, we suggest COCs as first-line therapy. We do not suggest one pill over another, as they are all effective. Other options are reviewed in detail separately. (See 'Hirsutism' above and "Management of hirsutism in premenopausal women".)

●Ovulation induction – For women who desire pregnancy, we suggest weight loss. (We suggest this to all women with PCOS who are overweight or obese.) If they are unable to lose weight or modest weight loss does not restore ovulatory cycles, we suggest ovulation induction with letrozole. (See 'Anovulatory infertility' above and "Treatment of polycystic ovary syndrome in adults", section on 'Women pursuing pregnancy'.)

●Pregnancy complications – We suggest against the use of metformin during pregnancy to prevent gestational diabetes mellitus (GDM) (Grade 2B). We also suggest against the routine use of metformin to prevent pregnancy loss in women with PCOS (Grade 2C). (See 'Prevention of pregnancy complications' above.)

●Management of metabolic issues – For women with PCOS who are found to have type 2 diabetes mellitus or impaired glucose tolerance (IGT), the management is the same as other non-PCOS patients with these disorders. (See 'Prevention of diabetes mellitus' above.)

Metformin is associated with an improvement in glucose levels and a reduction in blood pressure and low-density lipoprotein (LDL). In addition it may delay or prevent conversion of IGT to type 2 diabetes mellitus. (See 'Prevention of diabetes mellitus' above.)