Abstract
Diabetic men have benefited in the last 30 years from a significant improvement in total and cardiovascular mortality, whereas diabetic women have had no improvement at all. Moreover, recent research focused on the role of sex hormones in glucose homeostasis, and might account for different pathophysiologic mechanisms in the development of diabetes-related complications. Thus, care of diabetic women is a challenge that requires particular attention. The available data regarding gender-specific care of diabetes mellitus are uneven, rich in some domains but very poor in others. The large prospective trials performed in the last 20 years have assumed that the natural history of diabetes mellitus in men and women, as well as the efficiency of glucose-lowering therapies and management of hyperglycemic-related complications, could be attributable without distinction to men and women. We propose in this paper to analyze the published medical literature according to the specific management of diabetes mellitus in women, and to try to distinguish some particular features. We found important distinctions between diabetic men and women regarding the patterns of abnormalities of glucose regulation, epidemiology, development of diabetes-related complications, ischemic heart disease, morbidity and mortality, impact of cardiovascular risk factors, development of the metabolic syndrome, depression and osteoporosis, as well as the impact of lifestyle modifications or primary and secondary preventions on cardiovascular risk factors, and finally medical therapeutics. Moreover, special considerations were given to some particular aspects of the medical life in diabetic women, such as the features of gestational diabetes mellitus and the management of pregnancy in pregestational diabetic women, use of contraception, hormone-replacement therapy and polycystic ovary syndrome.
Keywords
Diabetes mellitus (DM) results from different pathophysiologic mechanisms, all leading to a dysfunction of glucose homeostasis. There are mainly two types of DM [1]. Type 1 DM may appear in childhood or early adulthood as an autoimmune disease, whereas Type 2 DM is a metabolic disease related to obesity and insulin resistance. However, this dichotomic classification does not include all the forms of DM, whose review is beyond the spectrum of the present work. Therefore, this review focuses on gender-specific care of Type 1 and Type 2 diabetic patients.
Gender-specific care of DM has not been particularly developed, but it is urgent to change this fact. Indeed, a recent study showed that between 1971 and 2000, diabetic men had a 43% relative reduction in the age-adjusted mortality rate (including cardiovascular mortality rate; this reduction is similar to that of nondiabetic men), whereas diabetic women had no reduction in total or cardiovascular mortality, and the all-cause mortality rate difference between diabetic and nondiabetic women more than doubled [2]. Moreover, recent research has focused on implications of sex hormones in glucose homeostasis, and the findings we will expose in the pathophysiology section suggest a gender specificity of molecular pathways involved in the development of insulin resistance; as a result, one can suppose that gender-specific mechanisms are involved in the development of diabetes-related complications, leading to gender-specific features of DM that could potentially open the way to gender-specific therapeutic possibilities.
We propose in this paper to analyze the published medical literature according to the specific management of DM in women, and to try to distinguish some particular features. For this purpose we screened MEDLINE (1980–2007) and Google Scholar for all the data related to the search terms: gender-specific treatment of diabetes mellitus, women and diabetes mellitus, prospective studies in diabetes mellitus, epidemiology of diabetes mellitus, therapeutic trials in diabetes mellitus, gender specificity of cardiovascular disease and diabetes mellitus, gender-specificity of metabolic syndrome. At this time there is only one review with evidence-based medicine recommendations that have been published [3] and few of the recommendations are of the highest quality of evidence.
Definition & diagnostic criteria of DM
The diagnosis of typical Type 1 DM is based on a cluster of different factors: younger age at presentation, clinical signs of hyperglycemia and insulinopenia, presence of specific antibodies and low C-peptide levels on blood tests [4]. The American Diabetes Association (ADA) published criteria for the diagnosis of Type 2 DM [5] and defined two prediabetic situations (
Gender specificity of results of diagnostic tests
There are marked gender differences in the results of FPG and OGTT: in a US study involving 4367 patients (25% women), fasting glucose (FG) and 2-h plasma glucose (PG) values in an OGTT were significantly lower and higher, respectively, in women than in men, independent of ethnicity, BMI and age [7]. This fact was corroborated in a European retrospective analysis of OGTT and FG values performed in 15,606 patients without a history of DM (7083 men and 8523 women) and 1325 patients with a history of DM [8]. Women had more elevated 2-h PG levels than men for each category of ages except for the fifth decade; the prevalence of isolated increased 2-h PG levels for the diagnosis of DM was significantly increased in women aged over 60 years than in men, whereas prevalence of isolated pathologic FG levels for the diagnosis of DM and IFG was significantly higher in men aged under 70 years than in women. In this study, increased incidence of diagnosis of DM in both elderly men and women was mostly related to increased 2-h PG rather than IFG. Hanefeld et al. published a prospective study involving 664 patients [9]. In men, there was a higher prevalence of IFG than in women (sex ratio: 1:4), while the latter demonstrated a significantly higher prevalence of IGT (sex ratio: 1:7). Men with IGT had higher levels of insulin resistance, whereas women with IGT had increased deficit of early and late insulin secretion, suggesting different pathophysiologic mechanisms for these different patterns of glucose homeostasis disorder. Rathmann et al. published a prospective study involving 1353 subjects (49% women) who completed an OGTT: 50% of newly diagnosed DM in women were diagnosed by isolated abnormal OGTT (only 34% in men) with normal FG [10]. Finally, in a study involving 3370 subjects with 54% women, prevalence of isolated elevated 2-h PG without IFG was slightly (without statistically significance) increased in women (1.9%) compared with men (1.4%) [11].
Definition and diagnostic criteria for diabetes mellitus in men and women
The three diagnostic criteria for diabetes mellitus, as defined by the American Diabetes Association in 1997, are (on two repeated occasions):
Fasting plasma glucose ≥7 mm/I (126 mg/dl)
Plasma glucose value 2 h after 75 g glucose ≥11.1 mm/I (200 mg/dl)
Casual glucose measurement ≥11.1 mm/I (200 mg/dl) in association with cardinal symptoms of diabetes (polyuria, polydipsia, weight loss)
The two prediabetic situations, as defined by the American Diabetes Association in 2003, are:
Impaired fasting glucose, defined as a fasting glucose value between 5.6 and 6.99 mm/I (100–125 mg/dl)
Impaired glucose tolerance, defined as a plasma glucose value 2 h after 75 g glucose between 7.8 and 11.09 mm/I (140–199 mg/dl)
The conclusion of these studies is that screening DM by the use of FPG would miss the diagnosis in many women as well as in older men, justifying the use of OGTT in these populations.
Gender specificity of risk factors for the development of DM
Risk factors for incidence of Type 2 DM differ in males and females [12]: elevated BMI, positive family history of DM and low high-density lipoprotein (HDL) were associated with increased incidence of DM in both genders; in men, high daily alcohol intake, regular smoking and elevated systolic blood pressure were associated with increased incidence of DM; in women, increased uric acid and physical inactivity were specifically associated with increased incidence of DM. In another study, risk factors for undiagnosed DM that might justify a screening test for DM, were hypertension and family history of DM in both men and women, but only hypertriglyceridemia in women and increased waist circumference in men were independently associated with DM [10].
Of note, one study involving 68,907 nurses demonstrated that obesity and physical inactivity were both independent risk factors for the incidence of Type 2 DM, but elevated BMI and waist circumference had a more pronounced impact than physical inactivity [13]. Relative risk of developing DM was 2.08 for lean women (BMI <25 kg/m2) with physical inactivity (exercise <2.1 metabolic equivalents of task h/week), 10.74 for active and obese women (BMI >30 kg/m2), and 16.75 for inactive and obese women.
Sex hormones & pathophysiology of glucose homeostasis disorder
Sex hormones and their specific receptors play a role in the incidence of DM.
Sex hormones
The prospective Massachusetts Male Aging Study, involving men aged 40–70 years, showed that low levels of testosterone and sex-hormone-binding globulin (SHBG) were associated with an increased risk of developing Type 2 DM [14]. The Rancho Bernardo Study showed that testosterone levels were inversely correlated with incidence of Type 2 DM in older men but directly correlated with Type 2 DM incidence and insulin resistance in women [15]. A systematic review of the literature regarding the link between Type 2 DM and endogenous sex hormones showed that [16]:
Diabetic men and women have lower and higher levels of testosterone, respectively, in comparison with age- and sex-matched nondiabetic individuals;
High levels of testosterone in men (449.6–605.2 ng/dl) are associated with a 42% decreased risk of developing Type 2 DM;
High levels of SHBG in women (>60 nmol/l) give an 80% lower risk of Type 2 DM, whereas high levels in men (>28.3 nm/l) only give a 52% risk reduction;
Low levels of SHBG are strongly correlated with Type 2 DM in women but not in men;
Estradiol levels are elevated in both men and postmenopausal women with Type 2 DM.
Low levels of estradiol in animal studies are associated with increased insulin resistance and adiposity in both males and females [17]. In humans, the role of estrogen in the regulation of glucose homeostasis remains controversial [18]. In males, the rare clinical situation of estrogen resistance via mutations in the estrogen receptor (ER) or the aromatase cytochrome P450 (involved in the transformation of androgens to estrogens) may be associated with the development of insulin resistance. In menopausal nondiabetic women, menopause is a risk factor for elevated FPG [19], and it is admitted that increased androgemestrogen ratio, abdominal fat, and development of insulin resistance associated with menopause are responsible for increased cardiovascular morbidity and mortality [20].
Sex-hormone receptors
Emerging data regarding the role of ER-α and -β on the expression of various genes, as well as clinical observations, led to consideration of their importance in glucose homeostasis [21]. ER-α and -β have different body distribution (uterus, vagina, ovaries, pituitary and mammary glands for ER-α; prostate, salivary glands, testis, vascular endothelium, smooth muscles, immune system, neurons for ER-β) and distinct biological functions; these receptors have antagonistic effects on the gene expression of GLUT4, the main protein involved in glucose uptake in skeletal muscle (and thus involved in the development of insulin resistance), with animal studies suggesting a repressive role of ER-β. Moreover, nongenomic activity of these receptors through interactions with transcription factors such as NFκB, specificity protein 1 and activating protein 1 can also modulate the expression of GLUT4 in an opposite manner, with ER-α-NFκB interaction avoiding repression of GLUT4 expression and ER-β inhibiting the expression of the GLUT4 though its interaction with specificity protein 1. Ratios of ER-α and -β in membrane cells determine the effects on GLUT4 expression and thus on insulin resistance in specific target organs; the use of ER-β antagonists might thus be a way to decrease insulin resistance and should be an area of research [21].
Moreover, ER may be involved in immunological reactions. Animal studies showed development of immune nephritis in ER-α−/- mice and myeloid leukemia in ER-β−/- mice [21]. The role of estradiol and ER in immune DM remains to be determined, but it was shown in both male and female mice that estradiol has antiapoptotic properties in cells, which are mediated by ER-α, including pancreatic β-cells, suggesting that estradiol has a protective effect against the occurrence of β-cell apoptosis via ER-α [22].
Sex hormones & sex-hormone receptors are responsible for the distribution of adipose tissues: implications for secretion of adipokines & insulin resistance
Sex-hormone receptors are present both in subcutaneous and visceral fat [23]. Androgen receptors are expressed more in visceral than subcutaneous fat, in both men and women, and are upregulated by testosterone [24]. Furthermore, ER-α upregulates the antilipolytic α-2A adrenergic receptor in subcutaneous fat but not in intra-abdominal adipose tissue [25]. These facts suggest a reason for a difference in adiposity patterns between men and women, with peripheral obesity (gluteal adiposity) in women and android obesity in men.
In one study, Pedersen et al. showed that ER-α is similarly expressed in subcutaneous gluteal adipose tissue, subcutaneous abdominal and intra-abdominal adipose tissue, but ER-β1 is five- and three-fold lower in the intra-abdominal adipose tissue of women and men, respectively, as compared with the subcutaneous adipose tissue. No significant gender differences in ER expression were detected in any of the fat depots investigated [26].
Moreover, the gender-specific obesity pattern influences the secretion of several adipokines by adipose tissues [27,28]. For example, leptin, which influences the central control of food intake, is in correlation with estrogen levels, is essentially produced by peripheral fat and is thus higher in women than in men, and low levels might be linked to higher central-fat obesity. Adiponectin, produced by adipocytes, presents insulin-sensitizing and antiatherosclerotic properties and is also usually higher in women than in men, although obese patients have reduced levels.
Impact of weight gain on the development of DM might differ between genders: a rate of weight gain more than 3 kg/year was assessed as a strong risk factor in all men but only in women who were not previously overweight. The best predictor for developing Type 2 DM is most probably a combination of BMI and waist circumference or waist:hip ratio in both sexes, but hazard ratios to develop Type 2 DM are higher in women than in men for each of these markers, underscoring the importance of reducing overweight and obesity in women [29].
Thus, women with normal gynoid repartition of adipose tissues have protective factors – that is, normal secretion of adipokines – against development of insulin resistance. The corollary of this is that women with abdominal obesity will have abnormal secretion of adipokines and thus increased insulin resistance, with increased risk of developing Type 2 DM.
Abdominal obesity, inflammation, endothelial dysfunction & cardiovascular morbidity
Abdominal fat tissue increases fatty acids and inflammatory cytokines, such as TNF-α and IL-6, which cause insulin resistance and cardiovascular adverse outcomes [28,30]. The role of inflammation in the pathophysiology of the incidence of Type 2 DM might be more important in women than in men, as C-reactive protein (CRP) levels (a key marker of inflammatory status) are significantly more elevated in women than in men diagnosed with DM [31].
The mechanisms by which inflammation leads to increased insulin resistance and cardiovascular mortality are beyond the scope of the present work, but it results in endothelial dysfunction, which is characterized by abnormal angiogenesis, blood flow and/or contractility [32], associated with the development of cardiovascular disease and an increased cardiovascular mortality. Of note, a small study involving Type 1 DM patients showed that the effects of oxidative stress were more marked in diabetic women than in men, as well as in the nondiabetic population, evoking the possibility that oxidative stress might aggravate the cardiovascular prognosis in diabetic women but not in diabetic men [33]. In addition, obesity and Type 2 DM in women might be associated with much more impaired endothelial dysfunction than men [34].
In conclusion, whereas both diabetic men and women have higher prevalence of central obesity, the impact on cardiovascular outcomes might be worse in diabetic women than men because abdominal obesity is associated in women with unusual metabolic profile in addition to a proinflammatory status, with worse endothelial dysfunction.
Epidemiology
Type 2 DM accounts for 85–95% of all diabetes in developed countries and even more so in developing countries. The prevalence of DM is directly correlated with the increasing age of the population and is higher in urban than in rural areas. The total number of women with DM is 10% higher than men, as well as the number of women with IGT, which is 20% higher than in men [301].
In the USA, data from the NHANES 1999–2002 indicated no significant difference in prevalence of diagnosed DM between men (6.7%) and women (6.3%), although the prevalence was higher in some minorities, such as non-Hispanic blacks and Mexican–Americans, and rose significantly with age (up to 21.6% in people aged over 65 years) [35]. The prevalence of undiagnosed DM was higher in men (3.6%) than in women (2.1%), especially those aged over 65 years. In Europe, prevalence of DM in patients aged over 30 years varies among countries, being less than 10% in patients aged under 60 years, except in Malta, and increases in both genders with older age; it is higher in men in the age group 40–59 years but is higher in women aged over 70 years, and is most elevated in women aged between 80 and 89 years, reaching numbers as high as 54.6, 54.2 and 56.5% in Oulu (Finland), Malta and Guia (Spain), respectively [8]. According to the role of estrogen in glucose homeostasis, as exposed in the prior section regarding sex hormone' impact on glucose homeostasis, one can hypothesize that menopause-related estrogen changes are the cause of increased insulin resistance and prevalence of DM in older women.
Estimated worldwide Type 1 DM prevalence in children aged under 14 years is 0.02%, with variation among countries and ethnicities [301]. A prospective study conducted in Europe for 2 years found a high geographic variability of Type 1 DM incidence from Slovakia (4.8/100000 person-years) to Leicestershire, UK (13.4/100000 person-years), with a 1:5 male:female ratio in the 25–29 years age group, and a higher ratio than 1 in the 14–25 years age group, although the ratio was 1:1 in the group aged 0–14 years (except in Slovakia, where there was a higher incidence of Type 1 DM in male children) [36].
Projections of overall prevalence of DM for year 2030, including Type 1 and Type 2 DM, show a doubling global number of individuals with DM (366 vs 171 million) mostly due to the increased aging and urbanization of the world population. The three countries with the highest numbers of persons with DM are, and will remain, India, China and USA [37].
In conclusion, the total number of women with DM is and will remain higher than the total number of diabetic men, although the prevalence of DM, which will increase in both genders, is and will remain slightly increased in men aged under 60 years, and slightly more elevated in women aged over 60 years [37]. In Type 1 DM, there is a 1:1 male:female ratio in patients aged 25 years or younger, but a 1:5 ratio in the group aged 25–29 years.
Complications of DM & related comorbidities
As we exposed the data related to the involvement of the sex hormones in the glucose homeostasis and the implications for developing more endothelial and vascular dysfunction in women in comparison with men, one might suppose that women are prone to more microvascular and macrovascular complications; thus, if this is true, one can also hypothesize that, at a same range of glucose disorder, the vascular dysfunction will be more marked in women than men. This is an area of future research, as recommended therapeutic target levels of glucose are, for now, similar in men and women [38].
Microvascular complications
Opthalmologic complications
Ocular complications related to DM are essentially cataracts and diabetic retinopathy (DR). The latter is considered to be the leading cause of blindness in people aged 20–75 years in developed countries [39]. We found only one publication that specifically assessed the impact of gender on DR in Type 1 DM patients [40]. DM was diagnosed before age 30 years; prevalence of DR in women was significantly higher than in men, but men had higher prevalence of severe DR. Prevalence of DR in women aged over 50 years increased with age (more in black women than white), whereas it decreased in men, and prevalence of clinically significant macular edema increased with age in all women (more in black than in white women), while not in men (not white or black). It seems there is no gender difference in prevalence of DR in Type 2 DM, and it is an independent predictor of mortality, especially cardiovascular, in both sexes; however, nonproliferative DR was more strongly associated with cardiovascular mortality in diabetic women than in men [41]; this finding is corroborated by three other studies [42–44]. Specific data regarding DR during pregnancy will be further discussed in the paragraph about the impact of pregnancy on DM.
Kidney complications
Diabetic nephropathy (DN) is defined by one of the following features: microalbuminuria, overt proteinuria or renal insufficiency, which may appear alone or in combinations during the course of the disease. It is the leading cause of end-stage renal disease requiring dialysis in developed countries and it is associated with an increased cardiovascular morbidity and mortality [45].
Gender impact on DN is a topic that has been reviewed in 2001 [46]. There are contradictory studies in both Type 1 and Type 2 DM-related nephropathy. However, female gender was never found to be associated with an increased incidence of DN or a worse nephrologic prognosis. Analysis of the sex-related influence on incidence and prevalence of DN is probably biased by other risk factors such as hypertension and length of diabetes. Opposite action of estrogen and testosterone on the kidney might explain gender differences in DN [3]. Indeed, estrogen reduces proliferation of mesangial cells and the activity of the renin–angiotensin–aldosterone system, whereas testosterone increases both. Moreover, estrogen increases the activity of metalloproteinase enzymes and NO synthesis, whereas it decreases collagen synthesis by mesangial cells. The effect of testosterone on the three latter systems is unknown. Finally, genetic factors might be associated with gender differences in the occurrence of DN. The M235T polymorphism in the angiotensinogen gene, for example, increases the incidence of DN in Type 2 DM male patients but not in female patients [47]. In addition, the AT2R gene seems to have an impact in the development of kidney dysfunction and hypertension in Type 1 DM male patients but not in Type 1 DM women [48]. These genetic factors should be further explored for a better understanding of their impact on gender difference in DN.
Diabetic neuropathy
In 2005, the ADA published a statement and review regarding diabetic neuropathies [49]. Although an epidemiological and therapeutic review was discussed in this statement with reference to large studies, there was no evidence of any gender specificity of the diabetic neuropathies. Some small studies considered gender as a risk factor for diabetic neuropathy. In a study involving 191 Type 2 DM patients (57% women), men had more distal symmetrical polyneuropathy and electromyography-supported neuropathy than women, but gender had no effect on the presence of neuropathic complaints [50]. In a case–control study involving 110 diabetic patients (78% women), male gender was associated with increased incidence of diabetic neuropathy [51].
According to autonomic diabetic neuropathy, we found one notable study that assessed a difference between men and women [52]. A metaanalysis that included 17 studies and involved 4584 patients (92% with Type 1 DM; 46% women) showed that at a given specificity of 86%, prolonged electrocardiographic QTc was 3.8-times more sensitive in men than in women to predict autonomic failure.
Macrovascular complications
Cardiovascular complications
There are emerging data today that ischemic heart disease differs between men and women, according to presenting symptoms, ways of diagnosis, evaluation and prognosis [53,54]. Yearly cardiovascular death affects more women than men, and there are no reduced cardiovascular-related mortality rates over the years in women as it is observed in men. An analysis of data published in the American population-based surveys (NHANES I-II-III) concerning the last 30 years reveals that diabetic men had a 43% reduction in the overall and cardiovascular mortality rates (similar to the rate observed in the nondiabetic population), whereas it did not change in diabetic women [2]. Several studies assessed the gender-related impact of DM on cardiovascular mortality and morbidity (
Recent studies regarding gender differences in cardiovascular mortality/morbidity in diabetes mellitus patients.
CHD: Coronary heart disease; CM: Cardiac mortality; DM: Diabetes mellitus; HR: Hazard ratio; MI: Myocardial infarction; OM: Overall mortality; RR: Risk ratio.
The increased cardiovascular morbidity and mortality in diabetic women may be due to several reasons [53]: women are less aggressively treated than men; women may have more combined cardiovascular risk factors than men; cardiovascular risk factors have worse impact in women than in men; cardiovascular risk factors are more severe in women than in men. The data related to cardiovascular risk factors in diabetic women will be discussed in the next section.
It is worrisome to note that diabetic women are less aggressively treated than men at the time of an acute coronary event [55]. Their management is characterized by less frequent use of aspirin, heparin or β-blockers; delayed used of thrombolytic therapy with greater incidence of major bleeding; and less frequent use of cardiac catheterization, percutaneous transluminal coronary angioplasty (PTCA) and coronary artery bypass graft surgery. Gender also influences outcomes of PTCA [56,57]; however, bad outcomes in diabetic women after PTCA were only found in old studies but not in recent ones, which showed an improvement in PTCA outcomes equivalent to the ones observed in diabetic men [58,59]. As some data in animals showed that local estrogen secretion might prevent in-stent restenosis [60], it raises the question of whether women benefit from estrogen-related protective effects on in-stent restenosis; there are no available data supporting this hypothesis. The use of drug-eluting stents also seems safe and efficient at 1-year of follow-up, with equivalent outcomes in men and women [61], although a recent meta-analysis showed that the use of a sirolimuseluting stent in diabetic patients might be associated with an increased hazard ratio for death in both sexes, warranting a tight follow-up after such a procedure in diabetic patients [62].
Finally, a different pathophysiology of cardiovascular disease between sexes may also worsen outcomes in diabetic women. In light of endogenous and exogenous hormonal wide impact, they might have a vascular dysfunction that might be more prevalent than in men and expressed independently of coronary obstruction [53]. Those differences between men and women may also be involved in the fact that diabetic women with heart failure have lower quality of life than men [63] and that they respond less well than men to standard therapy with β-blockers and angiotensin-converting-enzyme inhibitors in the treatment of heart failure [64].
In summary, as other authors [34,65,66], we consider that DM abolishes the protective estrogen-related benefit on CHD in nondiabetic women; diabetic women should be treated as aggressively as men in the occurrence of an acute coronary event.
Cardiovascular risk factors: gender differences to explain worse cardiovascular outcomes in diabetic women
In diabetic women, the odds ratio of each usual cardiovascular risk factor for adverse cardiovascular outcomes is higher than in diabetic men and the nondiabetic population [71]. Moreover, it appears that women are less aggressively treated, not only in the setting of an acute ischemic event but also in primary and secondary prevention, to reach recommended therapeutic goals of HbA1C, low-density lipoprotein (LDL) cholesterol levels or blood pressure [63,72]. Finally, women do not benefit similarly to men from the gold-standard therapies, such as statins or aspirin.
Diabetic dyslipidemia
Diabetic dyslipidemia (low HDL, hypertriglyceridemia and increased small LDL particles) seems to be more marked in diabetic women than in diabetic men, with more severe impact as a cardiovascular risk factor. Low HDL levels in DM women might be associated with a fourfold increase in CHD mortality compared with nondiabetic women or men with equivalent HDL levels; hypertriglyceridemia is more deleterious in women than in men [3,65,73]. The ADA recommends different therapeutic goals of HDL levels according to gender (50 mg/dl in women and 40 mg/dl in men), but recommends the same target LDL levels of lower than 100 mg/dl in both sexes [38].
Therapies for hyperlipidemia differ between men and women. First, a low-fat diet benefits men more than women with regards to the improvement of dyslipidemia and cardiovascular disease [74]. Second, it is not known if the use of elevating-HDL drugs might benefit women more than men as a means of prevention of cardiovascular disease, and this topic should be studied in randomized, controlled trials. Third, statins to reduce hyperlipidemia do not have similar effects in men and women. In the latter, they have no effect on CHD morbidity or mortality in primary prevention and on total mortality in secondary prevention [75]; in addition, they have fewer benefits on reduction of mortality and stroke events [76].
Hypertension
Hypertension seems to be more frequent in diabetic women than in diabetic men, with a more deleterious effect. An increase of 10 mmHg systolic blood pressure might lead to an increase of 30% CHD mortality in women but only 14% in men [3,73]. However, the ADA recommends the same blood pressure levels of less than 130/80 in both diabetic men and women [38]; this recommendation should be challenged and perhaps diabetic women will benefit from more aggressive treatment and lower targets for blood pressure. In the light of the particular role of inflammation in cardiovascular pathophysiology in women, it might be interesting to consider preferentially anthihypertensive drugs as renin–angiotensin–aldosterone-system blockers and β-blockers, which lower CRP levels [77].
Obesity
Obesity is a pandemic phenomenon, with an age-dependant increased prevalence, and increases more in women than in men [78]. Type 2 DM women might suffer a more negative impact of obesity and fat mass in comparison with men. They develop left ventricular hypertrophy more easily, which is a marker of cardiovascular mortality [79]; in addition, obesity and Type 2 DM might be associated with much more impaired endothelial dysfunction in women than in men [34]. In fact, elevated BMI is more strongly correlated with impaired cardiovascular risk factors (such as low HDL or high LDL levels, apolipoprotein A1, CRP) and adverse cardiovascular outcomes than inactivity in women [80]. Thus, lifestyle modifications in the form of exercise and diet to lose weight are dramatic issues in diabetic women. Exercise, even in lean women, has benefits with regards to cardiovascular outcomes [13].
Coronary calcification
Coronary calcification was correlated with severity of coronary stenosis in symptomatic Type 2 DM patients as well as in Type 1 DM patients of both genders. A study involving 302 Type 1 DM (55% women) found that Type 1 DM abolishes the protective effect of female gender on the prevalence of coronary calcification as seen in nondiabetic patients, although there was a lack of correlation between coronary calcification and CHD in diabetic women in comparison with diabetic men [81]. Another study involving Type 1 DM patients showed similar results; comparing prevalence of coronary calcification in nondiabetic women and men, there was a dramatic increase in Type 1 DM women but not in men [82].
Inflammation, oxidative stress & endothelial dysfunction
Uncontrolled DM and elevated HbA1C are correlated with increased CRP levels [83], which are associated with worse cardiovascular outcomes and increased cardiac mortality [84]. As we already noticed in the section regarding pathophysiology of glucose regulation and inflammation, the latter leads to insulin-resistance and endothelial dysfunction, which may be of higher range and associated with worse cardiovascular outcomes in Type 1 diabetic women [33] and in Type 2 obese diabetic women than matched diabetic men [34].
Metabolic syndrome
This syndrome has been distinctively defined by several organizations using different criteria of insulin resistance, glucose intolerance, dyslipidemia, obesity and hypertension [85–89]. In most countries there is an increase in the prevalence of the metabolic syndrome, which is higher in women than in men [78,90]. Individuals with the metabolic syndrome are at risk of developing Type 2 DM [91,92] and have an increased cardiovascular morbidity and mortality [93–96]. Women with metabolic syndrome may have a worse cardiovascular mobidity and mortality because of worse inflammation status determined by higher CRP levels than in men [97], more arterial stiffness [98], carotid stenosis [99], and a higher prevalence of ischemic strokes [100] than men with the same profile. Although we did not find studies that compared diabetic men with women with the metabolic syndrome, diabetic women with the metabolic syndrome should be considered as very-high-risk patients for cardiovascular disease, and screening and prevention or therapeutic considerations should be an area of future research in this population.
Sex-hormone status
Premenopausal women are considered to have a cardiovascular protective effect related to high estrogen levels. Development of DM might be associated with lower estrogen levels and increased levels of male sex hormones, thus worsening the cardiac clinical outcomes [66]. As exposed earlier, DM seems to abolish the protective effect of estrogens on cardiovascular disease in women.
Prothrombic status
Ulcerative coronary plaques are more frequently observed in patients with DM than in individuals without DM, probably owing to a procoagulation status, which is in part due to inflammation-related endothelial dysfunction, and partly due to hematologic changes associated with high-coagulable status and a lower platelet threshold activation, especially in women [101]. Aspirin is thus very important in diabetic patients, but has different effects in primary prevention according to gender. Women have less reduction of cardiac ischemic events but better preventive effects on stroke events than men [102], with more striking effects in the group aged over 65 years [103]. Effects of aspirin in secondary cardiovascular prevention are well known in men as well as in women [104]. The ADA recommends to start low-dose aspirin in any diabetic patient (Type 1 and 2) in both genders over the age of 40 years with additional risk factors (family history of CVD, hypertension, smoking, dyslipidemia or albuminuria) and to consider low-dose aspirin as primary prevention in diabetic patients aged 30–40 years with added risk factors, but not in young patients aged under 21 years [38,105]. Of note, reducing cardiovascular morbidity and mortality requires smoking cessation in diabetic men and women; in addition, alcohol consumption should be limited to one drink in women and two in men, daily [38].
Cerebrovascular complications
There is controversial literature regarding gender impact on stroke occurrence and prognosis. A subanalysis of the UKPDS study showed that female gender had a protective effect on the incidence of stroke in patients with Type 2 DM [106]. In a Canadian cohort study based on 61,327 individuals (55% women), there was a stronger association between stroke and diabetes in men than in women aged 45–64 years (odds ratio: 3.86 and 1.63, respectively), suggesting that diabetic men are more susceptible to stroke than diabetic women [107]. However, a British study that involved 41,799 diabetic patients and 202,733 control individuals showed a stroke rate of 11.9/1000 in diabetic patients, twice more than in the control group; the increase in risk attributable to diabetes was highest among young women (hazard ratio: 8.18; 95% confidence interval: 4.31–15.51) and decreased with age [108]. In a prospective study that involved 261 diabetic women and 300 diabetic men among 2446 patients hospitalized for ischemic stroke, diabetic women had a poorer prognosis, with an in-hospital mortality rate of 14.9 versus 8.3% in diabetic men [109]. In a study with Type 1 DM patients, incidence of stroke was more elevated in women than in men (26.1 and 17.9%, respectively), particularly in the age 40–49 years category, but also in other age subgroups; DN was a strong predictor of stroke [110].
These contrasting studies underscore the fact that the impact of gender on stroke in diabetic patients, as well as in the general population, should be further evaluated. Younger women with Type 1 or Type 2 DM might have more deleterious effects of glucose dysregulation than men with regards to ischemic stroke-related morbidity and mortality. This topic should be further studied in the future.
Foot complications
The term ‘diabetic foot’ includes a vast range of abnormalities, and involves macrovascular as well as microvascular diabetic complications. In a large, prospective study involving 500,848 diabetic patients and 500,268 matched subjects, women aged under 45 years had higher incidence of peripheral revascularization than men, and women aged over 84 years had higher incidence of both lower-extremity amputation and peripheral revascularization procedures than men [111]. There are few relevant publications that assess gender difference for this complication. A Swedish study showed that health concepts and beliefs, which differ between women and men, can lead to differences in preventing and managing diabetic foot lesions (women were more active than men in seeking and applying professional advice), thus influencing outcomes [112].
Depression
Diabetic patients have comorbid depression or psychological symptoms more often than healthy individuals. A meta-analysis showed that depression affects concerns up to a third of diabetic patients, and that the odds ratio for depression in Type 1 and Type 2 DM was two-times higher than in the nondiabetic population; the odds ratio was more elevated in diabetic women than men, corroborating the fact that depression has a female preponderance in the general population [113].
Depression in youth with Type 1 DM is up to three-times more prevalent than in nondiabetic youth; there are controversial data regarding gender difference in the occurrence of depression in young Type 1 diabetic patients [114]. A review article specifies a twofold higher risk of depression and psychological distress in diabetic women than in men, without distinguishing between Type 1 and Type 2 DM women [115]. In a study involving 53,072 individuals aged 20–64 years who had Type 1 or Type 2 DM, it was shown that DM was significantly associated with depression in women aged 20–39 years (odds ratio: 2.52), but not in older women [116]. The association was not significant in either age group in men, but it tended to be stronger in younger men. In another study, involving 825 Type 1 DM (48% women) and 786 Type 2 DM patients (50% women), there was a higher prevalence of anxiety and depression in women than men, as well as a stronger association of depressive symptoms with uncontrolled diabetes (high levels of serum HbA1C) in Type 1 and 2 diabetic women than in men; the authors of this study suggested that changes in estrogen levels might link glycemic control and depressive symptoms, and that women might be more negatively affected by depressive symptoms with regards to behavioral self-care treatment than men [117].
Impact of depression during the course of DM is very important [115,118]. It alters glycemic control and behavioral adherence to treatment, and may be associated with the development of diabetes-related complications; antidepressive medications might also lead to glycemic dysregulation via behavioral modifications and metabolic disorders. The gender difference of impact of depression in the diabetic population has not been studied, although depressive symptoms predicted cardiovascular events in Type 1 DM women but not in men [119].
We can conclude that women with DM, especially younger ones and those with uncontrolled DM, have an increased risk of depression in comparison with diabetic men. The pathophysiology of estrogen involvement on depressive symptoms and glycemic control is an area of research.
Osteoporosis
Bone metabolism is affected by DM. Diabetic osteopathy affects both men and women with either Type 1 or Type 2 DM. The assessment of bone mass is usually done with the measurement of bone mineral density (BMD) by a dual-energy X-ray absorptiometry exam. The clinical end point of osteoporosis is the occurrence of fracture.
BMD is lower in Type 1 diabetic individuals than in nondiabetic patients, both in adult men and women [120,121]. In a small retrospective study involving 60 Type 1 DM patients (50% women), men had significantly lower vertebral and femoral BMD values whereas women had only significantly lower femoral BMD in comparison with healthy age- and sex-matched patients [122]. In children, a prospective study showed a more markedly decreased bone mass in girls than in boys (after a mean duration of 6 years of the disease, and appearing at a mean age of 13.8 years compared with matched healthy individuals) [123]; however, this finding is controversial [124,125]. In Type 2 DM patients, BMD was similar [120] or higher [121,126] than in nondiabetic patients. In one study involving Type 2 diabetic patients, BMD values were higher in diabetic women and similar in diabetic men compared with nondiabetic individuals; the evocated reason was a probably higher hyperandogenism and hyperinsulinism in these women [127]. Another sex difference to note is a faster BMD loss in older (aged 70–79 years) white Type 2 DM women than in Type 2 DM men or black women; in the latter, there is no difference in the BMD loss compared with normoglycemic subjects [128].
The risk of hip fracture in women with Type 1 DM was evaluated to be 6.4-fold [129], 6.9-fold [130], 8.9-fold [131] and 12.25-fold [132] in comparison with nondiabetic women. Type 1 DM men have a 3.9-fold risk of nonvertebral fracture and a 17.8-fold relative risk of hip fracture in comparison with nondiabetic men [131], but another study did not find a significant difference in fracture rates between Type 1 DM men and nondiabetic men [130]. In Type 2 DM women, studies showed a 1.5-fold [130], 1.7-fold [132], 1.82-fold [133], twofold [131] and 2.2-fold [129] relative risk of hip fractures compared with normoglycemic women; of note, this increased risk of fracture was observed despite higher BMD in Type 2 DM women [133,134]. In Type 2 DM men, studies showed a less significantly but still more elevated risk of hip fracture [130,135]. Another study did not find a significant difference in risk of fractures in Type 2 DM men compared with nondiabetic men [131].
A special concern appeared recently regarding increased fractures and worsening osteoporosis in older women with Type 2 DM treated with thiazolidinediones. Shwartz et al. showed that in 69 out of 666 diabetic patients treated with thiazolidinediones (rosiglitazone, pioglitazone or troglitazone), women, but not men, had a yearly increased total, spine and trochanter bone loss in comparison with diabetic patients who did not receive one of these drugs [136]. A review of the safety data of the ADOPT study by GlaxoSmithKlein and an independent safety committee found that roziglitazone was associated with an increased incidence of fractures (hands, humerus and feet) in treated Type 2 DM women in comparison with men (9.3% in women vs 3.95% in men), whereas other tested oral antidiabetic drugs (metformin and glyburide) had no such effect on treated women in comparison with men [302]. In addition, Takeda pharmaceuticals, manufacturing pioglitazone, a concurrent peroxisome-proliferator-activated ᵧ-receptor agonist of rosiglitazone, published a letter announcing that the use of pioglitazone was associated with an increased incidence of fractures in women but not in men (1.9 fractures/100 treated patients vs 1.1 fractures/100 treated patients) in the same areas observed with roziglitazone: in the distal upper limb (forearm, hand and wrist) and the distal lower limb (foot, ankle, fibula and tibia) [303]. The information was diffused by the US FDA and it is now advised to consider the risk of fractures before starting these drugs in Type 2 DM women [304,305].
In summary, Type 1 DM women and men have, respectively, slightly and significantly reduced BMD levels compared with nondiabetic matched individuals, and both have more severe osteoporosis and more fractures than nondiabetic individuals, without straight marked gender difference. Type 2 DM women, while having higher BMD levels, have more fractures and accelerated bone loss than men. This point is emphasized by the recent concern of increased osteoporosis in older (aged 70–79 years) diabetic women treated with thiazolidinediones (roziglitazone, troglitazone or pioglitazone) but not in treated diabetic men. Modalities of osteoporosis screening, prevention and treatment in Type 1 diabetic patients and Type 2 diabetic women should be more investigated and are an area of future research.
Gynecologic & sex-hormone-related considerations in DM
Gender differences in puberty in patients with Type 1 DM
Young girls and female adolescents with Type 1 DM have puberty-related specificities compared with age-matched Type 1 DM male patients. A pubertal growth retardation was observed in Type 1 diabetic female patients with elevated HbA1C but not in young male patients [137]. Moreover, young Type 1 DM girls gain more weight than Type 1 DM boys [138–140] and have more difficulty losing weight than male patients [139]. Finally, glucose homeostasis is more difficult to obtain in girls with Type 1 DM at adolescence than in boys for several reasons: there might be a gender-difference sensitivity to insulin at puberty [141]; girls are more prone to discontinue insulin therapy to limit weight gain than boys [142]; girls have a higher incidence of eating disorders at puberty than boys [143]. It results that Type 1 DM girls may develop more microvascular complications than boys [144,145].
Menstrual abnormalities, infertility & sexual disorders in DM women
Irregular menstruation & infertility
In a retrospective study that compared 143 Type 1 diabetic women probands with their 186 healthy sisters and 158 unrelated nondiabetic controls, Type 1 DM was an independent risk factor for a report of any menstrual problem (including long cycle length, long menstruation, heavy menstruation) at younger age ranges (20–39 years) but not at older ages [146]. One conclusion of the authors was that menstrual abnormalities in the earlier reproductive years might be more accentuated in women with Type 1 DM.
In another analysis of this cohort study, Type 1 DM women also had delayed menarche and earlier menopause, which led to a 6-year reduction of reproductive period [147]. In a systemic review, up to 20% of Type 1 DM women had menstrual irregularities, 20% had hyperandrogenemia and 30% had hirsutism; these findings suggested a high prevalence of polycystic ovarian syndrome (PCOS) among Type 1 diabetic women, which was found to range from 12 to 40% [148]. Indeed, the diagnosis of PCOS [149–151] requests a combination of two out of three criteria: clinical and/or laboratory hyperandrogenism; oligo-anovulation; and morphologic ovary changes. A total of 50–70% of affected women have insulin resistance (in comparison with 10–25% of women in the general population) and thus have a five-to ten-fold greater risk of developing Type 2 DM. Up to 45% of women with PCOS will be diagnosed as having Type 2 DM or IGT [152]. The corollary of this is that Type 2 diabetic women are also at greater risk (five- to ten-times greater) of developing PCOS than nondiabetic matched cases [152]. Insulin resistance in women with PCOS is associated with an 11-fold increased risk of developing the metabolic syndrome with lipid-profile abnormalities in comparison with age-matched women without PCOS [153]. Thus, there is concern that PCOS is associated with an increased adverse cardiovascular morbidity and mortality. Of note, hyperandrogenism in PCOS might also increase the incidence of endometrial carcinoma (without any relation to the metabolic condition of affected women), is evidently associated with fertility disorder, and has implication for daughters and sisters who might also develop the syndrome and should thus also be screened [149].
Sexual disorders
In diabetic men, sexual dysfunction is essentially expressed by erectile trouble, which might result from endothelium and smooth muscle disorders as well as autonomic neuropathy and psychological causes; in diabetic women, sexual disorder is much less studied (pathophysiology, expression, risk factors and severity) [154]. In one review, sexual dysfunction in diabetic women, in contrast to diabetic men, was not related to older age, BMI, glycemic control, duration of diabetes or diabetes-related complications (except for chronic renal failure) [154]. Women with either Type 1 [155] or Type 2 [156] DM have a higher prevalence of sexual dysfunction than healthy women; in the former, it might only concern a disorder in the luteal menstrual phase [157]. In a questionnaire-based study involving 97 women and 95 men with Type 1 DM, up to 27% of women and 22% of men described sexual dysfunction, a higher percentage than in the general population; predictors of sexual dysfunction in men were both psychological (low-quality partnerships and poor emotional and cognitive adjustment to diabetes) and somatic factors (higher age, higher BMI, poor glycemic equilibration and diabetes-related complications), but were only psychological in Type 1 DM women (depression and poor cognitive adjustment to diabetes) [158]. A similar impact of depression on sexual disorder in diabetic men and women was also found in the previously cited review [154].
In conclusion, sexual dysfunction might radically differ between diabetic women and men according to the involved pathophysiologic pathways and risk factors, and should thus be further studied in the aim to better understand and prevent its occurrence.
Pregnancy & DM
Pregnant women may have pregestational (Type 1 or 2) DM (PGDM) or gestational DM (GDM), which is defined as DM diagnosed for the first time during pregnancy.
Complications of maternal hyperglycemia concern both mothers (hypertension, polyhydramnios, pre-eclampsia, infections and so on) and fetuses; reactive fetal hyperinsulinism may lead to macrosomia with complicated delivery (dystocia) and need for cesarean surgery, neonatal hypokalemia, hypocalcemia, polycythemia, hyperbilirubinemia or hypoglycemia [159]; spontaneous abortions and congenital malformations are more frequent in Type 1 and 2 PGDM than in GDM [160,161]. Except for the rare sacral dysgenesis and femoral hypoplasia unusual facies syndrome, there is no diabetes-related specific malformation, but a trend to more elevated rates of cardiovascular, neurologic, gastrointestinal, genito-urinary and skeletal anomalies in PGDM women's newborns exists [162,163]. Glucose homeostasis is critical until the end of organogenesis, at the seventh week of gestation, to avoid congenital abnormalities and spontaneous abortions [164]. Thus, diagnosis and treatment of GDM, as well as optimal glucose regulation in PGDM, are challenging and critical issues.
GDM
Pathophysiology
Pregnancy is a particular situation with marked modifications in glucose metabolism [165]. The first trimester of pregnancy is characterized by an increased insulin sensitivity especially due to an increase in first-phase insulin secretion after nutrient stimulation; progressively, hepatic gluconeogenesis is increased to supply fetal nutrition, but a maternal peripheral insulin resistance develops because of hormonal (estrogen and progesterone) effects on lipogenesis, secretion of placental diabetogenic hormones such as human sommatomammotropin, increased levels of endogenic Cortisol and glucagon, and pregnancy-related diabetogenic effects of prolactin. A normal pancreatic function counteracts these diabetogenic changes, but GDM may develop in case of adaptation failure [165].
Epidemiology
Incidence of GDM varies with ethnicity and increases with maternal age from 0.4% in Caucasian women aged under 25 years to 5.5% in women aged over 25 years, but up to 16% in Native American women [159]. Known risk factors are: age over 25 years (or 30 years for some authors); ethnicity (any non-Caucasian); first-degree relative with Type 1 or 2 DM, or GDM history; BMI above 25 (27 for some authors); previous history of GDM; own maternal birth weight below 2.5 kg or tenth percentile; a previous macrosomic baby; a previous bad outcome of pregnancy (stillbirth); polyhydramnios and/or a suspected large-for-gestational-age fetus in the current pregnancy [159].
Screening & diagnostic tests
Screening for GDM is a growing public-health concern because, as we noted in the section regarding obesity and the metabolic syndrome, prevalence of these two strong risk factors for developing Type 2 DM increases among young women [78,90]; thus, there will be a growing number of women with pregestational insulin resistance, which can lead to overt DM during pregnancy.
Extensive reviews exist on controversies about modalities and cutoff values regarding the screening and diagnostic tests for GDM [159,163,166]. Screening of the described risk factors is not sensitive enough and should not be the only method used [159]. A usual screening test is based on a 50-g OGTT with cutoff values of either 1-h glucose greater than 130 or 140 mg/dl according to the ADA [166]. Alternatively, some authors proposed to screen for GDM by using glucose levels [159]. If FPG or postprandial plasma glucose (PPG) meet criteria for the diagnosis of DM, there is no need to perform any further test; based on the fact that glucose levels are lower in the first half of pregnancy, some authors recommend a screening FPG of less than 85 mg/dl (4.8 mm/l), while others suggest less than 65 mg/dl (3.9 mm/l), before considering a diagnostic test. If the screening test is positive, then a diagnostic test should be performed. In the USA, a 100-g OGTT with two abnormal values is interpreted according to two widely used thresholds (PFG >95 mg/dl, 1-h glucose >180 mg/dl, 2-h glucose >155, 3-h glucose >140 mg/dl [167], or PFG >105 mg/dl, 1-h glucose >190 mg/dl, 2-h glucose > 165 mg/dl, 3 h glucose >145 mg/dl [168]), but the WHO recommends a 75-g OGTT with threshold values of FPG greater than 126 mg/dl and 2-h glucose greater than 140 mg/dl [85].
Management of GDM
Management of GDM [159,160,166] initially requires a low-carbohydrate diet (30–40% carbohydrates of the total daily caloric intake into three meals with 3–4 snacks, or lowering daily glycemic index, with 60% of carbohydrate) and physical activity; institution of pharmacologic therapy usually depends on fasting and PPG thresholds (FG of 95 mg/dl or PPG 155 mg/dl after 1 h or 130 mg/dl after 2 h for the ADA [166], but FG of 105 mg/dl and 140 mg/dl after 1 h and 120 mg/dl after 2 h for other authors [160]). Some authors studied the decision management of starting insulin therapy during GDM based on abdominal fetal circumference of greater than the 75th percentile, with FG >120 mg/dl or 2-h PPG >200mg/dl. It appeared to be a reasonable approach in comparison with conventional management based on glycemic criteria alone [169]. The usual recommended treatment is an intensive insulin therapy with the aim of maintaining normoglycemic status, but in recent years, several studies showed interest in oral hypoglycemic agents [170]. This issue will be discussed later, in the section regarding therapeutic considerations in diabetic pregnant women.
Delivery should be performed with a cesarean surgery if the fetus is macrosomic with an evaluated weight above 4555 g, whereas this recommendation is not warranted in the case of an evaluated fetal weight of 4000–4500 g [171]. Early labor induction at 38–39th week of gestation is an admitted policy in GDM women who are pharmacologically treated, with a diminution of macrosomic-related complications of delivery and stillbirth, although these advantages have not been found in cases of diet-treated GDM women [171].
Recurrence & overt DM following GDM
Rate of recurrence of DGM was evaluated in a meta-analysis to range from 30 to 84% in subsequent pregnancies [172]. Of note, 10–50% of women with GDM are diagnosed with DM within 5 years of the delivery, and thus screening with 50-g OGTT at 6 weeks after delivery is recommended [38,163,173]. If negative, subsequent evaluation is recommended within 3 years, but controversies exist regarding whether it should be based on triannual OGTT (perhaps the most cost effective) or annual FPG, or even HbA1C [173].
Discussion: who should be screened & what is the benefit:risk ratio of screening, diagnosing & treating GDM?
There are diverging considerations of the screening and diagnostic testing of GDM, because there are controversies regarding the benefit of treating GDM after it is diagnosed. Indeed, a Cochrane meta-analysis published in 2003 failed to find suitable trials to show evidence that pharmacologic therapies in GDM ameliorate outcomes, and did not find significant effect in treating pregnancy-related IGT except for limiting neonatal hypoglycemia [174]. However, a study published in 2005 showed that untreated GDM may lead to a 59% increase in global neonatal morbidity with up to fourfold increase in metabolic/macrosomia adverse events, whereas treated GDM is associated with an 18% increase in global neonatal morbidity without difference in comparison with nondiabetic pregnant women with regards to metabolic/macrosomia complications [175]. This may explain why screening tests are systematically recommended to all pregnant women in 94 and 84% of obstetric units in the USA and the Canada, respectively, whereas in the UK only 16% of obstetric units recommend systematic screening (72% recommend a diagnostic test in case of risk factors) [159]. However, screening only women with the described risk factors may miss up to 43% of cases of GDM [176].
Hollander et al. discuss whether, considering the lack of evident benefit of treating GDM, we are doing more harm than good in screening and diagnosing GDM [159]: 1.4 and 4% of women reported vomiting during a 75-g and 100-g OGTT, respectively; strict diet may lead to ketosis and alter fetal psychomotor development; and there are more cesarean procedures and more health costs. Moreover, tight glycemic control in diabetic women may be associated with hypoglycemic events, which can have deleterious effects for the fetuses, as observed in Type 1 PGDM mothers: 40% described at least one severe event during pregnancy, and the nutritional deficiency leading to hypoinsulinism is associated with intrauterine growth failure. The latter may be further responsible for increased neonatal mortality, cognitive deficits, increased incidence of diabetes, coronary artery disease, and hypertension at adulthood [177].
However, a meta-analysis showed that various different screening tests might identify various degrees of gestational hyperglycemia, but because of a lack of good clinical trials, no screening test appeared to be better than another, and thresholds at which health outcomes start to be altered remain unclear; as a result, the impact of treatment might also differ among different women with GDM according to the severity of glucose homeostasis disorder, and this domain should be actively explored [178].
Conclusion
GDM is intuitively associated with an increased morbidity and mortality for fetuses and mothers, but prospective, controlled trials are lacking to define the exact modalities of screening and diagnostic protocols, as well as the exact benefits of treatment. As long as these data are missing, gynecologists and physicians following pregnant women should apply one of the locally admitted protocols to all pregnant women, in the aim to treat hyperglycemia in good time and not to later regret a noninterventionist position. This policy is not accepted by all the organizations, and is based on expert opinions only; further studies are needed to validate this opinion or alternatively to define best specific criteria for population screening, thus limiting the number of pregnant women to be screened. Moreover, studies defining the best glucose thresholds requiring therapeutic interventions will be welcomed. Finally, postdelivery screening for DM also remains a critical issue.
PGDM
Proportions of Type 1 and Type 2 PGDM differ by ethnicity; for example, in France, a third of all PGDM is due to Type 2 DM and two-thirds are due to Type 1 DM, but in the Mexican–American population, 90 and 10% are due to Type 2 and Type 1 DM, respectively [163]. However, as there is a growing epidemic of obesity in younger adults, there is growing incidence of Type 2 DM in young women and, thus, Type 2 PGDM is becoming a growing concern [161].
Outcomes of pregnancy
A Danish national study, which included 990 women with Type 1 PGDM, reported perinatal morbidity and congenital malformations in comparison with outcomes in healthy women's pregnancies [179]. The perinatal mortality rate was 3.1% in Type 1 PGDM (risk ratio [RR]: 4.1), and the stillbirth rate was 2.1% (RR: 4.7). The congenital malformation rate was 5.0% (RR: 1.7). Better outcomes, with less perinatal mortality and congenital malformations, and lower cesarean surgery and preterm delivery rates, were observed in women who performed daily self-monitoring of blood glucose, as they had lower HbA1c values than women who did not measure their daily profile. However, even at normal HbA1C levels, risk of congenital malformation in newborns of Type 1 PGDM women is higher than in the general population [180], probably because of intermittent hyperglycemia [181,182]. During a 12-year follow-up of 182 women with Type 2 PGDM, pregnancy outcomes were assessed: fetuses had a twofold greater risk of stillbirth, and infants had a 2.5-fold greater risk of a perinatal mortality, a 3.5-fold greater risk of death within the first month and a sixfold greater risk of death at up to 1 year [183]. They had an 11-times greater risk of congenital malformations. Compared outcomes of pregnancy in Type 1 and 2 PGDM were studied in a large prospective British study that involved 2359 PGDM women (1707 Type 1 PGDM and 652 Type 2 PGDM) [184]. Similar rates of perinatal mortality (mean: 32/1000 pregnancies) and congenital malformations (mean: 46/1000 pregnancies) were observed; the former was fourfold increased and the latter was twofold more elevated than in the nondiabetic population. A smaller study, with 330 Type 1 PGDM, 540 Type 2 PGDM and 325 GDM patients, which persisted during the postpartum period, found a nonsignificant difference in the rate of pregnancy loss in Type 1 and Type 2 diabetes (2.6 and 3.7%, respectively), although the causes of pregnancy loss differed [185]. In Type 1 PGDM, more than 75% were due to major congenital anomalies or prematurity, whereas in Type 2 diabetes, more than 75% were attributable to stillbirth or chorioamnionitis. Women with Type 2 and Type 1 diabetes had similar HbA1c at presentation and near term, but the former were older and more obese and thus may have significant adverse factors for pregnancy loss.
Impact of pregnancy on DM
Type 1 DM women are prone to microvascular complications, which can worsen during pregnancy [162,164]. Pregestational renal failure reduces the chance of successful pregnancy, and the more elevated the creatinine levels are before conception, the greater the risks of worsening kidney function during pregnancy and after delivery, especially if pregestational proteinuria exists; this may further decompensate hypertension and lead to pre-eclampsia; uncontrolled hypertension might aggravate pregestational DR, which can worsen spontaneously in direct correlation with the pregestational severity of the retinopathy, but also with glucose levels before conception and the magnitude of improvement in glucose homeostasis in the first trimester [164]. However, pregnancy does not seem to influence the long-term prevalence or the severity of peripheral diabetic neuropathy, although a mild aggravation may occur in the short term (less than 2 years after delivery). A worrisome pregestational complication is gastroparesis, which is associated with up to twofold perinatal mortality, congenital malformations, polyhydramnios, preterm delivery, pre-eclampsia, hypoglycemic episodes and ketoacidosis [162].
There are much fewer specific data regarding pregnancy in Type 2 DM, but both diabetes-related maternal complications (nephropathy, retinopathy and also pre-eclampsia) and perinatal morbidity and mortality are identical to those observed in Type 1 PGDM [161].
Preconceptional care
Preconceptional evaluation can help to equilibrate glycemic control, treat DR and equilibrate hypertension with change of angiotensin-converting-enzyme inhibitors, angiotensin receptor blockers and β-blockers to antihypertensive drugs adapted to pregnancy [186]; thus, morbidity and mortality of both mothers and fetuses will be reduced [162,187]. Ray et al. found, in a meta-analysis that included 14 studies and 1292 pregnant women (most of whom were Type 1 PGDM), that preconceptional care could reduce the rate of congenital abnormalities from 6.5 to 2.5% [188]. For both Type 1 and 2 PGDM, a good glycemic control at least 3 months before starting pregnancy is recommended to achieve normoglycemia during fertilization and organogenesis; this requires the learning of self-monitoring of glucose in Type 2 DM women who did not check glucose levels previously, dietetic consultation and physical activity [187].
PGDM management
Both Type 1 and 2 PGDM pregnant women need a good follow-up of retinal evaluation and urinary protein range as well as blood pressure values to ensure that they do not develop diabetic complications during pregnancy, especially if pregestational complications are known [187].
The management of Type 1, as well as insulin-treated Type 2, PGDM requires insulin dose adaptation. The needs for insulin differ among pregnant PGDM women and vary during the pregnancy, as well as varying from one pregnancy to another in the same patient. In one study, which described pregnancy outcomes in 107 Type 1 PGDM women, need for insulin was stable until the 20th week, and then raised significantly until the 30th week of pregnancy, to reach up to 60% increase of the insulin dose from the baseline regimen [189]. However, a reduction in needs of insulin doses has been described between the seventh and 12th week of gestation in Type 1 PGDM pregnant women [190]. The global trend of increasing needs of insulin until the delivery, is due to worsening insulin resistance related to physiologic changes occurring during pregnancy, as described earlier [164,191]. A usual regimen increases from 0.7 U/kg/day in three to four doses during the first trimester up to 1 U/kg/day at the time of delivery; obese women may need more elevated doses of insulin. In Type 1 PGDM, delivery should be programmed for the 38th week of gestation to avoid stillbirth; this may necessitate early induction of labor and cesarean surgery, but minimizes the size of the fetus and macrosomic-related complications of delivery [171]. Whatever the Type of PGDM, normal glycemia at time of delivery is fundamental, and should not excess 8 mm/l because of risk of neonatal reactive hypoglycemia in the newborn [189]. Thus, maternal glucose levels between 70 and 90 mg/dl are recommended by the use of subcutaneous or even intravenous insulin injections, keeping in mind that with the delivery, there is a brisk reduction in insulin-resistance, which might necessitate giving of glucose with the continuation of insulin therapy, and to decrease insulin doses to prepregnancy ranges [164].
Type 2 DM is usually treated before the pregnancy with oral hypoglycemic agents in conjunction with low-caloric diet and exercise. It is admitted to switch from oral hypoglycemic agents to intensive insulin therapy as soon as the diagnosis of pregnancy has been performed; this allows optimal glucose control and minimizes teratogenic drug effects [163]. However, as noted earlier, safety and efficacy of oral hypoglycemic agents during pregnancy has been studied in recent years, and will be further defined in the pharmacologic considerations of glucose regulation during pregnancy. The same considerations for glucose normalization as in Type 1 PGDM are required at the time of delivery, as well as considerations for cesarean surgery according to macrosomic features of the fetus [171].
Conclusion
Type 1 and 2 PGDM are similarly associated with perinatal mortality and congenital malformations, the latter being more increased than in GDM. A preconceptional evaluation starting 3 months before conception seems to limit the adverse outcomes, and may also reduce the negative effect of pregnancy on pregestational diabetic complications such as retinopathy and nephropathy. An optimal glucose equilibration is fundamental from the time of organogenesis until the seventh week of gestation, but also later to minimize bad outcomes. Timing and mode of delivery should be based on fetal macrosomic evaluation, and may require early labor induction and/or cesarean surgery.
Contraception in diabetic women
The planning of pregnancy in diabetic women is crucial to avoid maternal as well as fetal morbidity and mortality that may result from prolonged pregestational hyperglycemia; however, hormonal contraception in diabetic women is of dramatic concern because of the fear of adverse effects such as disturbance in carbohydrate and lipid metabolism, and thrombotic events [192,193].
A systematic Cochrane review regarding hormonal and nonhormonal ways of contraception in Type 1 and 2 diabetic women found only three suitable randomized, controlled trials [193]. The authors concluded there was insufficient evidence to assess whether progestin-only and combined contraceptives differ from non-hormonal contraceptives in glucose and lipid profiles as well as in long-term complications; however, the safest way of contraception in diabetic women appears to be the copper intrauterine device (IUD), and low-dose estrogen combined oral contraceptives (COCs) might have advantages in young women without vascular complications; progestin-IUD might also be safe, and progestin-only contraception might be appropriate in women with microvascular or macrovascular complications.
A prospective study that was published after the Cochrane review, involving 55 Type 2 DM and 58 Type 1 DM perimenopausal women, assessed the metabolic impact of different regimens of COCs (ethinylestradiol [EE] 20 mg/desogestrel 150 mg, EE 30 mg/desogestrel 150 mg and EE 30 mg/gestodene 75 mg) and levonorgestrel-releasing or copper IUD as means of contraception [194]. The authors concluded that no regimen negatively influenced the diabetic equilibration (assessed by a stable HbA1C), but Type 1 DM women who received a combination of 30 mg EE/75 mg gestodene needed to significantly increase the dose of their insulin therapy; only the combination of 20 mg EE/150 mg DSC in perimenopausal women with Type 1 DM had a negative impact on the lipid profile within 6 months, which was mostly correlated with the equilibration of the DM, with an amelioration within 12 months; they noted a slightly increased platelet activity with a homeostatic disorder (but without clinical expression) in women who received COCs, which was milder with the 20 mg EE/150 mg DSC regimen, in comparison with neutral effects of IUD.
Of note, women with previous GDM should not be treated with progestin-only contraceptive therapy because it is associated with an increased incidence of Type 2 DM [195,196].
During the menstruation period, some women might present changes in their eating patterns and decreased insulin sensitivity, which require tighter blood control during this period without any relation to contraceptive regimen; periodic diabetic ketoacidosis was described in women with DM, but also hypoglycemic events, and the exact hormonal mechanism by which glucose homeostasis is affected remains unknown [197]. Of note, postmenopausal women treated in the past with hormonal contraception might have a lesser degree of subsequent ischemic heart disease independently of other risk factors, including DM [198].
In conclusion, there is not enough evidence to assess the dangers of hormonal contraception in diabetic women; the safest contraceptive method is the use of the copper IUD, but low-dose estrogen COCs might be suitable in young diabetic women without any vascular complications, whereas progestogen-releasing IUDs or progestin-only oral contraceptives might be safe in women with microvascular or macrovascular complications.
Postmenopausal status & hormone-replacement therapy in diabetic women
Hormone-replacement therapy (HRT) is a controversial option for postmenopausal women. On one hand, it helps controlling symptoms related to postmenopausal status and prevents osteoporotic bone fracture; on the other hand, it is associated with increased incidence of uterine and breast cancer, uterine bleeding, and cardiovascular and thromboembolic events [199]. These potential adverse outcomes are of serious concern in diabetic women, who are already at increased risk of cardiovascular events, especially during menopause, which is associated with worsening components of the metabolic syndrome: increased androgen:estrogen ratio, increased abdominal fat, and development of insulin resistance [20]. Moreover, Type 1 DM might be associated with increased incidence of endometrial cancer [200], and thus HRT might further aggravate this occurrence in these patients.
A meta-analysis that included 107 studies suggested that nondiabetic women who received HRT had reduced androgenic fat, insulin resistance, new onset of diabetes, blood pressure, LDL:HDL ratio and procoagulant markers; diabetic women had only reduced fasting glucose and insulin resistance [201]. In one large, randomized, multicenter, placebo-controlled trial [202], which involved 2763 postmenopausal women (734 with DM and 218 with IFG) with coronary heart disease, participants received combined conjugated estrogen (0.625 mg) and medroxyprogesterone (2.5 mg) or placebo and had a follow-up of 4.1 years. In all women, including the diabetic ones, there was a reduction of weight, waist circumference and waist:hip ratio; in the normoglycemic women, a 35% reduction of DM incidence was observed in comparison with matched women who received a placebo.
In the meta-analysis, oral HRT (conjugated rather than esterified estrogen, and estrogen alone rather than combined with progestin) had better metabolic effects than transdermal agents, but also led, in a dose-dependent association, to increased triglycerides and CRP levels and decreased protein S levels, whereas transdermal agents did not [201]. The clinical impact of these laboratory changes was not assessed. The authors concluded that younger postmenopausal women, including diabetic women, might benefit from transdermal HRT without increase of cardiovascular risks, whereas risks of oral HRT outweigh benefit, especially in older women with cardiovascular disease.
As in nondiabetic women, the potential positive metabolic effects of HRT cannot be translated in straightforward positive clinical cardiovascular effects. One retrospective study involving 6017 women with Type 2 DM aged 45–80 years showed that estrogen use independently of progestin combination led to a decreased incidence of cardiovascular events [203]. Another large, prospective study involving 24,420 diabetic women without recent MI showed a protective effect of low-dose estrogens (less than 0.625 mg estradiol) but not of higher-dose estrogens, which was more marked when combined with progestogens after 1 year of treatment but not during the first year of treatment [204]. Diabetic women with recent MI had an increased risk of recurrent ischemic event when treated with HRT. Finally, Howard et al. found a worsening atherosclerosis progression in diabetic and prediabetic women treated with different HRT regimens in comparison with healthy postmenopausal women, despite amelioration of components of the metabolic syndrome, probably due to elevation of CRP and fibrinogen levels [205]. As a result, it is today admitted that one should not start HRT for cardiovascular primary or secondary prevention in postmenopausal women, including recently postmenopausal women, because the latter might benefit from other cardiovascular prevention means such as statins and aspirin [199,206,207]. Other therapeutic alternatives to HRT include bisphopshonates to control osteoporosis, and selective serotonin-reuptake inhibitors such as venlafaxin or antiepileptics such as gabapentin to control vasomotor symptoms related to estrogen deficiency [208].
Some authors consider that HRT could be given in low-cardiovascular-risk postmenopausal diabetic women (without other contraindications such as breast or endometrial cancer) according to a stratification based on age, triglyceride levels and time from beginning of menopausal symptoms to starting HRT [208]. Short-term, low-dose, transdermal estrogen therapy (rather than long-term, high-dose oral therapy) in younger postmenopausal women seems to be the safest HRT regimen to control menopausal symptoms with regards to cardiovascular disease, but in very-low-risk diabetic postmenopausal women, one could also consider an oral HRT regimen. However, as we previously reviewed the worsening outcomes of cardiovascular disease in diabetic women, we think that DM excludes the possibility of a low-cardiovascular-risk category in affected women.
In conclusion, a complex relationship exists between method of administration, dose and type of HRT regimens and metabolic effects in postmenopausal women. Despite an amelioration of many components of the metabolic syndrome, there is no assurance of protective cardiovascular effects, especially in oral estrogen regimens that elevate CRP and triglyceride levels. Thus, considering that DM confers a high-risk status to postmenopausal women with regards to cardiovascular disease, we recommend to try alternative therapies instead of HRT as first line; if the latter is given, it should be transdermal, at low dose and for a short period of time. Further studies are needed for this major issue, which will involve more and more women as the combination of an aging population and increased incidence of DM and ischemic heart disease in the elderly will elevate the number of women concerned.
Pharmacologic considerations for glucose regulation in diabetic women
Glucose-lowering therapy
Drugs used to lower blood glucose may be oral hypoglycemic agents (e.g., biguanide, sulfonylurea, thiazolidinedione, α-glucosidase inhibitors), new oral dipeptidyl peptidase inhibitors, inhaled insulin, subcutaneous analogue or human insulin, subcutaneous glucagon-like peptide 1; for each of them, therapeutic trials have been conducted, but out of the epidemiologic data specifying the number of male and female patients included, we did not find any attempt to analyze the results by gender. The benefits of lowering hyperglycemia on prevention of microvascular complications have been similarly attributed to men and women, but there is no trial that aimed to check a gender difference in the effects of glucose-lowering therapy. However, we showed that there are emerging data suggesting gender difference in the pathophysiology (inflammation and endothelial dysfunction have a more significant role in women than in men) as well as in the course of DM. As we already stated, a meta-analysis showed that diabetic men but not women benefited of improved all-cause and cardiovascular mortality in the last 30 years [2]. Therapeutic trials should assess these factors and perhaps we will discover that some medications are more adapted to women than other ones. Perhaps the usual hypoglycemic drugs are not as efficient in women as in men? Perhaps they have secondary effects in women but not in men? Perhaps they are as efficient in men and women to reach specific cutoff target values of glucose, whereas the latter should perhaps be lower in women than men? These interrogations raise the more absolute question of whether management of DM should be different between men and women, with different objectives for fasting and PPG levels, for example. A gender analysis of the clinical studies would help us to understand better the causes of gender differences in mortality in diabetic patients.
In light of the role of the ER in glucose homeostasis, some authors consider specific ER-β antagonists and specific ER-α agonists as good options for DM equilibrations in women [21]; these possibilities should be carefully evaluated. Moreover, in light of the more important role of inflammation in the pathophysiology of diabetes and cardiovascular complications in women than men, one can hypothesize that thiazolidinediones, which appear to reduce CRP levels better than insulin [77], may be a candidate as first-line therapy in women? These questions should be urgently assessed in well-designed studies.
According to the gender-specific adverse side effects of antidiabetic therapies, as we already noted, a study showed that the use of thiazolidinediones was associated with more fractures in diabetic women than men [136]. This would counterbalance the benefits of reducing CRP levels; it shows the complexity of the pharmacologic hypoglycemic options, and underscores the fact that any hypoglycemic drug should be assessed according to advantages and disadvantages on glycemic, cardiovascular and global effects in women, rather than according to one isolated parameter of amelioration.
Glucose-lowering therapy during pregnancy
Insulin is the most widely used therapy to control glucose levels in pregnancy complicated with DM (GDM, Type 1 and 2 PGDM); however, increasing interest for oral hypoglycemic agents during pregnancy for GDM and Type 2 PGDM might lead to changes in the usual considerations of hypoglycemic therapies during pregnancy.
Insulin
Type 1 PGDM is treated with intensive insulin therapy before the beginning of the pregnancy; Type 2 PGDM women treated with oral hypoglycemic agents are usually switched to insulin therapy with two to four daily injections as soon as pregnancy is diagnosed, and GDM is usually treated with insulin after failure of diet and exercise to obtain normal glucose levels for pregnancy, as explained earlier.
Until recently, only human nonimmunogenic insulins were authorized during pregnancy [209]. With the rapidly increasing use of analogue insulins in the general nonpregnant diabetic population, dramatic benefits appeared in comparison with human insulin. When compared with human regular insulin, rapid-analogue insulins such as lispro and aspart reduce rates of hypoglycemia and control PPG levels better, with a diminution of HbA1C in Type 1 and 2 DM [210]; the newer rapid-analogue insulin glulisine has comparable effects [211]. Long-active analogue insulins such as glargine and detemir better control basal glucose as well as FG and PPG levels and reduce rates of hypoglycemia [210]. These advantages could be significant during pregnancy considering the hyperglycemia-related morbidity and mortality for both mothers and fetuses, as well as considering the possible adverse outcomes of hypoglycemia during pregnancy, as described earlier [177].
The question of safety regarding the use of analogue insulins during pregnancy is based on their potential modification of IgF1 levels, which is implicated during early pregnancy in the implantation of the embryo in the endometrium, and later regulates fetal growth as a mediator of the human placental growth hormone [191]. Owing to structural identity between the insulin receptor and the IGF-1 receptor, there are cross-reactions between natural insulin and the IGF-1 receptor, as well as IGF-1 with the insulin receptor; they activate distinct signaling pathways, a metabolic one for the insulin receptor and a mitogenic one for the IGF-1 receptor [212]. Thus, an insulin analogue, which has structural modifications owing to changes in amino acid sequences in comparison with human insulin, may have enhanced or reduced affinity for the insulin receptor and IGF-1 receptor, and might alter the natural processes mediated by IGF-1 [213]. Particularly, in vitro studies on animal and human cell lines showed that some experimental insulins with high affinity for the IGF-1 receptor had a correlated increase in their mitogenic activity and tumorigenic potency, and thus it should be recommended to use an insulin analogue with minimum IGF-1 activity [191]. The long-acting analogue insulins glargine and detemir have in vitro increased and decreased affinity for the IGF-1 receptor [214]. However, there are no data currently available on the use of detemir during pregnancy, and the use of glargine during pregnancy was reported only in 13 Type 1 PGDM and nine GDM; although there were no reported congenital abnormalities, these data are too poor to conclude about safety and efficacy of its use in pregnancy [191]. The rapid-analogue insulins aspart and lispro seem to have similar and slightly increased affinity for the IGF-1 receptor, respectively, compared with the human rapid insulin [214]. In total, 795 cases of use of lispro and 59 cases of the use of aspart during pregnancy were described in the literature [191], showing safety (no placental crossing; no more congenital abnormalities) and efficacy (as good as and maybe more efficacious than human regular insulin) compared with human regular insulin, with a more friendly user profile as time separating the injection from the meal is reduced [160,191]. However, concern regarding the use of lispro during pregnancy was raised when three women under a lispro regimen with Type 1 PGDM without pregestational DR developed proliferative retinopathy and needed laser therapy during pregnancy [215]. However, a propsective trial with 69 Type 1 PGDM women did not find a difference in the progression of retinopathy in women treated with lispro compared with those treated with human insulin, and showed better glucose control [216].
Oral hypoglycemic agents
In GDM, in case of failure of dietetic and exercise measures to reach satisfactory glucose levels, a pharmacologic therapy is started based on considerations exposed earlier in the paragraph regarding GDM. Intensive insulin therapy is usually started, but interest for oral hypoglycemic agents has been increasing in the last few years [170].
Glyburide is an oral sulfonylurea hypoglycemic agent that does not cross the placenta and has no known teratogenicity; its use in pregnancy achieved good glycemic control, similar to that obtained under insulinotherapy and without difference in rates of any complication [160,170,217–220]. Other oral hypoglycemic agents, in particular metformin, a class B medication during pregnancy, could also be efficiently used, especially with regards to the particular insulin resistance occurring in the pathophysiology of GDM discussed earlier, but was not studied enough in clinical trials [170]. It seems to cross the placental barrier and may be associated with more elevated rates of pre-eclampsia [221], but reassuring data were reported in one study with 32 Type 2 PGDM women who had started metformin before the gestation and continued it until delivery [221]. Although these women had more factors predicting adverse outcomes (e.g., higher BMI, HbA1C, blood pressure), there was no difference in morbidity in comparison with women who were not treated with metformin with regards to perinatal loss, early delivery and respiratory distress. Moreover, a study on safety and efficacy of metformin in PCOS was published with data regarding 39 women with PCOS who became pregnant under metformin (at a dose of 2.55 g in 80% of cases and 1.5 g in 20% of cases) treatment and pursued it until delivery, with a significant benefit in terms of reduced GDM rates with no drug-related complications in comparison with PCOS women who did not receive metformin during pregnancy; the preconceptional benefit effects of metformin were marked for reduction of BMI and insulin resistance [222]. Metformin might also be safe during lactation [223]. Thiazolidinediones are classified as class C drugs for pregnancy, can cross the placenta and were not studied specifically during pregnancy; their use is not recommended unless specific advantages would result [170].
In summary, insulin therapy remains the treatment of choice to control DM during pregnancy, with the best evidence of safety and efficacy for the use of human insulin. The use of the rapid-analogue insulin lispro seems safe for the fetus and efficient for glycemic control, but caution should be raised with regards to the potential development of DR in women who receive this insulin. The rapid-analogue insulin aspart seems safe for the fetus and efficient for glycemic control, but poor clinical data are available according to its use during pregnancy. There are not enough available data to recommend the use of long-active analogue insulins, but detevir seems safer than glargine with regards to the lower affinity for the IGF-1 receptor of the former, whereas the latter has a higher affinity. This is a particular area of research. Finally, Type 2 PGDM and GDM may, in the future, be managed with oral hypoglycemic agents. The best-studied agent with good evidence of safety and efficacy is glyburide, but the use of metformin should be more urgently explored, as it suits perfectly the particular pathophysiology of glucose dysregulation during pregnancy by diminishing insulin resistance. A large prospective trial with different therapeutic regimens (insulin, glyburide, metformin) should be performed before this approach is generalized.
Conclusion & future perspective
The available data regarding gender-specific care of DM are uneven. The large prospective trials performed in the last 20 years have assumed that efficiency of glucose-lowering therapies, as well as management of hyperglycemic-related complications, could be attributable without distinction to men and women, whereas a much higher number of men than women were included in these trials and there was no gender-specific analysis of the results. However, today, there is a growing awareness that gender influences semeiology, ways of diagnosis, prognosis and therapeutic efficiency in all the medical domains. In this review, we exposed the critical points that are obvious today with regards to the specific management of diabetic women, and we underscored the ones that remain to be explored in the future (
Areas of interest for clinical and basic science future research
Role of sex hormones in microvascular and macrovascular complications of diabetes mellitus (DM).
Impact on the occurrence of diabetic nephropathy in men and women of genetic (the M235T polymorphism in the angiotensinogen gene and the AT2R gene) and hormonal (estrogen and testosterone) factors should be further explored.
Pathophysiology of estrogen involvement on depressive symptoms and glycemic control.
Screening and management of osteoporosis in diabetic women.
Best screening and diagnostic tests of gestational DM.
Best therapeutic options and strategy to manage gestational DM, particularly the role of oral hypoglycemic agents.
Best strategy of screening for Type 2 DM after delivery in women with gestational DM.
Safest contraceptive regimen for diabetic women should be evaluated in prospective clinical trials.
The link between insulin resistance and polycystic ovarian syndrome, as well as the impact of insulin therapy on the development of the latter, should be further explored.
Clinical trials checking if more aggressive therapy in diabetic women with regards to both glycemic and blood pressure therapeutic targets can reduce mortality in diabetic women, and thus challenge the idea that the same therapeutic targets benefit women and men similarly.
The role of ER-β antagonists and ER-α agonists in glucose homeostasis in women.
The role of high-density lipoprotein-elevating agents in diabetic women with regards to effects on cardiovascular morbidity and mortality.
Safety of different analogue insulins during pregnancy and further evaluation of consequences of their interactions with the IGF-1 receptor.
Executive summary
Gender specificity of results of diagnostic tests for diabetes mellitus (DM): screening of DM in women and older men should include an oral glucose tolerance test and not only be based on fasting glucose values.
Gender specificity of risk factors for DM: elevated uric acid levels and physical inactivity are specific risk factors to develop DM in women, whereas alcohol intake, smoking and systolic hypertension are specific to men. In both genders, risk factors to develop DM are: family history of DM, elevated BMI or waist circumference, and low high-density lipoprotein values.
Sex-hormones: testosterone has gender-dimorphic effects on the incidence of Type 2 DM. Elevated levels are protective in men but worse in women; low levels are associated with insulin resistance in men. Estrogen deficiency in men and postmenopausal women may be associated with increased insulin resistance.
Sex-hormone receptors: estrogen receptors (ER)-α and -β have opposite effects on glucose homeostasis. The former favors glucose homeostasis through direct effects on GLUT4 transcription; the latter downregulates the expression of GLUT4. The membrane ER-α:β ratio determines the global effects on GLUT4 expression.
Sex hormones and sex-hormone receptors are responsible for the adipose tissue distribution. Women with gynoid repartition of fat tissues benefit from the protective effects of a normal pattern of secretion of adipokines, such as leptin and adiponectin, against development of insulin resistance. These protective effects are diminished when women develop android obesity.
Abdominal obesity, inflammation, endothelial dysfunction and cardiovascular morbidity: abdominal obesity is associated with increased secretion of proinflammatory cytokines such as TNF-α, which impair normal glucose homeostasis and lead to endothelial dysfunction, which is more marked in diabetic women than men, and may play a key role in the increased female cardiovascular morbidity and mortality.
The absolute number of women with Type 2 DM and impaired glucose tolerance is 10 and 20%, respectively, higher than in men.
Ophtalmologic complications: prevalence of diabetic retinopathy is increased in Type 1 diabetic women but is more severe in men aged under 50 years. Nonproliferative diabetic retinopathy is more strongly associated with increased cardiovascular mortality in Type 2 diabetic women than men.
Kidney complications: some studies show increased prevalence and incidence of diabetic nephropathy in men. Sex hormones as well as angiotensinogen and AT2R genes may play a role.
Diabetic neuropathy: small studies suggest increased incidence of distal symmetrical polyneuropathy in diabetic men in comparison with women, but only very poor data are available. This topic should be further explored. Prolonged QTc is more sensitive in diabetic men than women to diagnose cardiac autonomic neuropathy.
Cardiovascular complications: DM abolished estrogen-related protective effects on coronary heart disease in women, and thus leads to higher mortality rates than in men, especially in female patients with long duration of DM.
Cardiovascular risk factors: diabetic women have worse levels of cardiovascular risk factors than men, with stronger deleterious effects.
Cerebrovascular complications: morbidity and mortality might be worse in women, particularly young diabetic women, in comparison with men.
Incidence of depression is more elevated in young diabetic women than men.
Type 1 diabetic patients of both genders have lower bone mineral densitometry levels and more elevated rates of osteoporotic fractures. Type 2 diabetic women have higher bone mineral densitometry levels but more elevated rates of osteoporotic fractures in comparison with nondiabetic women and diabetic men.
Puberty in Type 1 diabetic patients: glycemic equilibration is more difficult to achieve in girls than boys and may thus develop more microvascular complications.
Menstrual abnormalities, infertility and sexual disorders in diabetic women:
– Up to 40% of Type 1 diabetic women will develop features of polycystic ovarian syndrome, and Type 2 diabetic women have up to tenfold increased risk to develop this syndrome.
– Sexual dysfunction in diabetic women, in contrast to diabetic men, is mostly due to psychological factors (depression and poor cognitive adjustment to DM).
Gestational DM:
– Controversies exist to define the best screening and diagnostic strategies.
– Usual management of glycemic control is based on insulin therapy but there is growing interest for oral hypoglycemic agents such as glyburide and metformin.
– Women with gestational DM are at increased risk to develop Type 2 DM. The best screening strategy remains to be defined.
Pregestational DM:
– Type 1 and 2 diabetic women have similar adverse outcomes of pregnancy with increased perinatal mortality, stillbirth rates and congenital malformations.
– Preconceptional care of diabetic patients is critical to assess and treat diabetic microvascular complications that can worsen during pregnancy, as well as to achieve good glycemic control before the conception to minimize risks of congenital malformations.
– Pregnant women should be regularly assessed for development of diabetic retinopathy during pregnancy, especially if treated with analogue insulins.
Contraception in diabetic patients:
– Intrauterine devices have no metabolic or thrombotic effects in comparison with hormonal contraceptives, and are thus safer to use in diabetic patients.
– Low-dose estrogen combined oral contraceptives might be a suitable option in some diabetic patients.
Postmenopausal status and hormone-replacement therapy in diabetic patients: hormone-replacement therapy should not be prescribed in diabetic patients before other alternative therapies have been considered.
– There are no available data today showing that efficacy of any hypoglycemic agent differs between men and women to achieve glycemic control.
– Thiazolidinediones seem to lead to more fractures in women than men.
– Diabetes management during pregnancy:
– Human rapid regular insulin and delayed neutral protamine Hagedorn are safe and efficient regimens to equilibrate DM during pregnancy. They are the treatment of choice for glycemic control during pregnancy.
– Analogue rapid- and long-acting insulins may have advantages over human insulins in reducing events of hypoglycemia and better controlling glucose levels. However, analogue insulins interact with the IGF-1 receptor in a different manner than human insulins, and thus modify mitogenic function of IGF-1. Use of analogue insulins with low affinity for the IGF-1 receptor is safer.
– Oral antidiabetic agents such as glyburide and metformin are of particular interest for the treatment of gestational DM and pregnant women with Type 2 DM, but should be further evaluated in clinical trials.
Footnotes
The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.