| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Departments of Endocrinology (H.F.E.-M., J.S., H.d.l.C., R.G.-R.) and Pediatric Endocrinology (B.R., R.B., M.A.), Hospital Ramón y Cajal, Madrid, Spain 28034
Address all correspondence and requests for reprints to: Héctor F. Escobar-Morreale, M.D., Ph.D., Department of Endocrinology, Hospital Ramón y Cajal, Carretera de Colmenar km. 9.100, 28034 Madrid, Spain. E-mail: hector.escobar{at}uam.es
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Eighty-five women with type 1 diabetes mellitus were evaluated for symptoms and signs of hyperandrogenism. In 68 of the patients, several serum androgen and hormone concentrations were measured. The polycystic ovary syndrome (PCOS) was defined by the presence of menstrual dysfunction, together with clinical and/or biochemical evidence of hyperandrogenism, and exclusion of other etiologies. Eighteen healthy women, menstruating regularly, served as controls for the androgenic profiles.
Thirty-three patients (38.8%) presented hyperandrogenic disorders (16 had PCOS, and 17 had hirsutism without menstrual dysfunction). Type 1 diabetic patients with PCOS presented increased serum total and free testosterone concentrations, and serum androstenedione levels, but had normal serum sex hormone-binding globulin and dehydroepiandrosterone-sulfate levels. Hirsute type 1 diabetic women without menstrual dysfunction presented normal serum androgen levels. There were no significant differences between hyperandrogenic and nonhyperandrogenic type 1 diabetes mellitus women in clinical variables such as the duration of diabetes, age at diagnosis of diabetes, conventional or intensive insulin therapy, mean daily insulin dosage, or metabolic control.
In conclusion, women with type 1 diabetes mellitus have a high prevalence of hyperandrogenic disorders, including PCOS and hirsutism.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Insulin has a synergistic stimulatory effect with LH on the ovarian production of androgens (5), and this effect seems to be mediated both by a direct action on the ovarian insulin receptor and by a complementary action on the intraovarian insulin-like growth factor system (4). Moreover, the amelioration of the increased circulating insulin concentrations in women with PCOS, in response to weight loss, diazoxide, metformin or troglitazone, has been shown to reduce free testosterone concentrations and to increase sex hormone-binding globulin (SHBG) (6, 7, 8, 9). These findings point to a central role of insulin in the androgen metabolism of these patients.
Type 1 diabetes mellitus is a disease characterized by an autoimmune injury to the endocrine pancreas, resulting in the complete abolition of endogenous insulin secretion. Under insulin treatment, however, type 1 diabetes mellitus patients may also have insulin resistance, especially when obesity is also present (10).
The current recommendation is to maintain patients with type 1 diabetes mellitus in a strict metabolic control by means of intensive conventional insulin therapy, and supraphysiological doses of insulin are usually needed to achieve this aim (11). Thus, a certain degree of hyperinsulinism may be present in these patients, resulting from insulin resistance and from the relatively excessive insulin doses employed in their treatment, and this hyperinsulinism might increase the adrenal and ovarian androgen secretion, as occurs in women with type 2 diabetes mellitus. To our knowledge, the prevalence of hyperandrogenic disorders in women with type 1 diabetes mellitus has not been reported previously.
In the present study, we have evaluated the prevalence of PCOS, as defined by the criteria derived from the conference sponsored in 1990 by the National Institute of Child Health and Human Development [ovulatory dysfunction, clinical, and/or biochemical evidence of hyperandrogenism, and exclusion of other etiologies such as hyperprolactinemia and congenital adrenal hyperplasia (12)], as well as the prevalence of hirsutism, in the population of women with type 1 diabetes mellitus followed up in our Hospital, focusing on possible predisposing factors related to the evolution and treatment of type 1 diabetes mellitus.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Women with type 1 diabetes mellitus. All the postmenarcheal women diagnosed with type 1 diabetes mellitus corresponding to the population assigned to our Hospital by the Spanish National Health System were proposed to participate in the study. Of the whole population of 99 postmenarcheal women with type 1 diabetes mellitus, 11 patients refused to enter the study, and 3 were excluded (one who was pregnant at the time of the study, and 2 who had concurrent untreated hypothyroidism and hyperprolactinemia, conditions which might induce hypertrichosis or an increase in adrenal androgens).
Therefore, 85 women with type 1 diabetes mellitus [age, 22.7 ± (SD) 5.3 yr; body mass index, 23.2 ± (SD) 2.7 kg/m2] were finally studied. A detailed clinical history was obtained, including age at the time of the study, age at diagnosis of type 1 diabetes mellitus, age of menarche, type of insulin treatment employed (conventional vs. intensive), mean insulin doses, and indexes of metabolic control during evolution of type 1 diabetes mellitus.
A complete physical examination was performed. Hirsutism was defined by the presence of excessive body hair distributed in an androgen-dependent pattern, with a modified Ferriman-Gallwey score (13) of 8 or more, which was determined by a single experienced observer. Menstrual disturbances were evaluated on recall for every patient and control. Oligomenorrhea was defined by 3 or more cycles with a length of more than 36 days in the previous year, and amenorrhea was defined by lack of vaginal bleeding for the last 3 months (14). The presence of acne was also evaluated and was considered to be a sign of hyperandrogenism only when persisting after the second decade of life.
Controls. Eighteen regularly menstruating nondiabetic women [age, 26.2 ± (SD) 3.8 yr; body mass index, 22.9 ± (SD) 3.6 kg/m2], without signs or symptoms of hyperandrogenism or family history of endocrine diseases, served as healthy controls for the basal steroid plasma concentrations.
All the women included in the study were Caucasian. None of the patients or controls had received sex steroid medications, including oral contraceptives, for the last 6 months. Informed consent was obtained from every patient and control. The study was conducted according to the Guidelines of the Institutional Review Board of the Hospital Ramón y Cajal and to principles expressed in the Declaration of Helsinki.
Analytical procedures
Studies were performed during the follicular phase, between days 5 and 10 of the menstrual cycle, or in amenorrhea after excluding gestation by appropriate testing, both in patients and in controls. Sixty-eight of the 85 patients with type 1 diabetes mellitus and the 18 healthy controls reported to the Endocrine-Metabolic testing room between 0800 and 0900 h, after a 12-h overnight fasting. An indwelling iv line was placed in a forearm vein, and, after 15–30 min, basal blood samples were obtained for measurement of total testosterone (T), SHBG, LH, FSH, estradiol, androstenedione, 17-hydroxyprogesterone (17-OHP), and dehydroepiandrosterone-sulfate (DHEAS). The normal ranges for serum androgen and steroid precursors were calculated from the results of the healthy women.
Samples were analyzed as previously described (15, 16, 17). The free testosterone concentration (FT) was calculated from T and SHBG concentrations, assuming a serum albumin concentration of 43 g/L, and taking a value of 1 x 109 L/mol for the association constant of SHBG for T and a value of 3.6 x 104 L/mol for that of albumin for T (18). In the women with PCOS, serum PRL concentrations were measured, in a single assay, using an automated immunochemiluminescence assay (Immulite, Diagnostic Products, Los Angeles, CA, with a mean intraassay coefficient of variation of 6.2%.
Clinical variables in type 1 diabetic women
Several clinical variables that might be related to the development of hyperandrogenism in diabetic patients were obtained from the patients’ records. Retrospective analysis of the time from onset of type 1 diabetes mellitus, age at the onset of type 1 diabetes mellitus, age of menarche, and number of patients with onset of type 1 diabetes mellitus before menarche, were compared among type 1 diabetic women with or without hyperandrogenism.
The mean number of daily insulin doses and the mean daily insulin dose since diagnosis of type 1 diabetes mellitus were calculated from the registered doses recorded every 3 months. The mean daily perimenarcheal insulin dose was calculated for the period beginning 2 yr before menarche and ending 2 yr after menarche. The mean body mass index since diagnosis of type 1 diabetes mellitus was calculated from the data measured every 12 months, and the mean hemoglobin A1c levels since diagnosis of type 1 diabetes mellitus were calculated from data obtained at 3- or 6-month intervals for each patient. The differences among hyperandrogenic and nonhyperandrogenic type 1 diabetic patients in these variables were also evaluated.
Definition of PCOS in the study population
The diagnosis of PCOS was based on endocrine criteria, according to the conference sponsored by the National Institute of Child Health and Human Development in 1990 (12). Although a consensus was never reached, most of the attendants to this conference agreed that the diagnosis of PCOS requires the presence of ovulatory dysfunction, clinical and/or biochemical evidence of hyperandrogenism, and exclusion of other etiologies such as hyperprolactinemia, congenital adrenal hyperplasia, and androgen-producing tumors (12). However, no indication of how to define ovulatory dysfunction, hirsutism, or hyperandrogenemia was given.
As in previous studies regarding the prevalence of PCOS in the general population (19, 20), we have considered menstrual dysfunction (oligomenorrhea or amenorrhea) as indicative of ovulatory dysfunction. Hirsutism (defined by a modified Ferriman-Gallwey score of 8 or more) and/or acne persisting after the second decade of life were considered clinical signs of hyperandrogenism. Serum T, FT, androstenedione and/or DHEAS concentrations above the 95th percentile of the control group of healthy women defined biochemical hyperandrogenism.
Finally, in women matching the first two criteria for the diagnosis of PCOS, hyperprolactinemia, nonclassic congenital hyperplasia, and androgen-secreting neoplasms were excluded. Hyperprolactinemia was excluded when serum PRL levels were less than 24 µg/L. Nonclassic adrenal hyperplasia was ruled out by a basal serum 17-OHP concentration less than 6 nmol/L [which is the cut-off level suggested for maximum sensitivity using 17-OHP measurements obtained from early-morning samples (21)]. In two patients presenting with a basal 17-OHP level more than 6 nmol/L, the diagnosis of nonclassic adrenal hyperplasia was excluded by an ACTH- stimulated serum 17-OHP level less than 30 nmol/L (22). None of the subjects included in the study had serum testosterone levels above 6.9 nmol/L and/or serum DHEAS levels higher than 21.7 nmol/L, usually considered as the limits above which an androgen-secreting tumor must be suspected (23).
Statistical analysis
Data are represented as means ± SD in the text
and in Table 1
and as means ±
SEM in Fig. 1. For continuous
variables, the mean values of the different groups of diabetic patients
and the group of healthy controls were compared by the one-way ANOVA,
followed by the protected least-significant-difference test for
multiple comparisons. The homogeneity of the variances was evaluated by
Levene’s test. Logarithmic or square-root transformations were applied
before ANOVA, to ensure homogeneity of the variances, as needed. In
women with type 1 diabetes mellitus, the comparisons of the clinical
variables between women with PCOS, hirsute women without menstrual
dysfunction, and nonhyperandrogenic women, were made by one-way ANOVA
or Pearson’s 
2 tests, as appropriate.
P < 0.05 was considered as statistically
significant.
|
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
When including the whole population of women with type 1 diabetes mellitus, irrespective of whether serum samples were available at the time of the study, 33 type 1 diabetic women (38.8%) were considered to have hyperandrogenism [16 patients (18.8%) matched the 3 criteria for the definition of PCOS, and 17 patients (20%) presented hirsutism with regular menstrual cycles].
The diagnosis of PCOS was based on the presence of hirsutism and menstrual dysfunction, with or without increased serum androgens, in 9 women; on the association of menstrual dysfunction and increased serum androgens in 6 subjects; and on the presence of acne after the second decade of life, hyperandrogenemia, and menstrual dysfunction in 1 diabetic patient. In the 15 women diagnosed with PCOS in whom serum samples were obtained, serum PRL and basal or ACTH-stimulated 17-OHP levels ruled out hyperprolactinemia and congenital adrenal hyperplasia, respectively.
Another 17 (20%) type 1 diabetic women presented hirsutism without menstrual dysfunction, with or without increased serum androgens. Thus, hirsutism was present in 26 diabetic patients (30.6%), including 9 of the women presenting with PCOS. Their hirsutism score ranged from 8–19, with mean ± SD of 11.6 ± 3.3. Acne persisting after the second decade of life was found in 1 type 1 diabetic woman with PCOS. None of the women with type 1 diabetes mellitus presented male pattern balding, alopecia, or acantosis nigricans.
We also calculated these prevalences including only the women who were tested for serum hormone levels, and we obtained similar results. Of a total of 68 patients, 15 presented with PCOS (22.1%) and 13 presented had hirsutism without menstrual dysfunction, with or without increased serum androgens (19.1%), for a global prevalence of hyperandrogenic disorders of 41.2%.
Comparison of the hormone profiles among the different groups of diabetic patients, and healthy controls
Subjects were classified into four groups: type 1 diabetic
patients with PCOS, hirsute type 1 diabetic women without menstrual
dysfunction, type 1 diabetic patients without clinical evidence of
hyperandrogenic disorders, and healthy controls. No differences in age
and body mass index were found among the three groups of patients with
type 1 diabetes mellitus (Table 1
).
Serum total and free testosterone levels, and serum androstenedione
concentrations, were increased in type 1 diabetic patients with PCOS,
compared with healthy controls and nonhyperandrogenic type 1 diabetic
women. (Fig. 1). Hirsute type 1 diabetic women without menstrual
dysfunction presented intermediate serum concentrations of total and
free testosterone, and of androstenedione, that were not different,
compared with any other group (Fig. 1).
No differences among the groups were observed in the serum SHBG, DHEAS,
17-OHP, LH, FSH, and estradiol levels (Fig. 1).
Clinical variables in type 1 diabetic women
The groups of type 1 diabetic women with PCOS or hirsutism without
menstrual dysfunction, and nonhyperandrogenic diabetic women, were not
different in terms of time from onset of type 1 diabetes mellitus, age
at the onset of type 1 diabetes mellitus, age of menarche, number of
daily insulin doses, mean daily perimenarcheal insulin dose, mean daily
insulin dose since diagnosis of type 1 diabetes mellitus, mean body
mass index since diagnosis of type 1 diabetes mellitus, and mean
hemoglobin A1c levels since diagnosis of type 1
diabetes mellitus (Table 1). Eighty-eight percent of the patients were
on long-term intensive insulin treatment at the time of the study,
receiving three or four daily doses. There was no statistically
significant difference among the three groups of diabetic patients in
the number of patients with onset of diabetes before menarche (Table 1).
However, when considering, as a whole, the group of type 1 diabetic
patients with PCOS, together with hirsute type 1 diabetic women without
menstrual dysfunction, these patients showed a near-significant trend
toward a higher number of subjects with onset of diabetes before
menarche, compared with nonhyperandrogenic type 1 diabetic women (64%
vs. 42%, 2 = 3.67,
P = 0.055).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The 18.8% prevalence of PCOS in women with type 1 diabetes mellitus is
considerably higher (2 = 7.4,
P < 0.005) than the 6.5% prevalence of PCOS we have
recently found in a group of 154 unselected nondiabetic Caucasian women
from Madrid, Spain, which is the city from which come the women with
type 1 diabetes included in the present study (20).
Moreover, a 4.7% prevalence has been recently reported by Knochenhauer
et al. (19) in unselected white women from
Alabama, and a 6.8% prevalence has been found in the Greek island of
Lesbos (25). These results suggest that the prevalence of
PCOS is similar in nondiabetic women with different geographic and
ethnic backgrounds.
Hirsutism is a marker of androgen excess at the pilosebaceous unit and
is considered as a clinical sign of hyperandrogenism, even in the
absence of menstrual dysfunction and/or increased serum androgen
concentrations. We have found a 30.6% prevalence of hirsutism in type
1 diabetic women, as defined by a hirsutism score of 8 or more,
considering as a whole the hirsute type 1 diabetic patients with PCOS,
and hirsute type 1 diabetic patients without menstrual dysfunction.
This figure is markedly increased (2 = 21.3,
P < 0.0001), compared with the 7.1% prevalence of
hirsutism, defined by a hirsutism score of 8 or more that we have found
in nondiabetic women from Madrid (20).
Knochenhauer et al. (19) found hirsutism, defined by a modified hirsutism score of 6 or more, in 8% of unselected white women, and this figure was further reduced to 2.8% if a hirsutism score of 8 was used as the cut-off value. As reviewed by Knochenhauer et al. (19), older studies have reported excess hairiness in 5–15% of consecutive Caucasian women (26, 27, 28, 29, 30), and Ferriman and Gallwey (31) found only 5% of the general population had a score of 6 or more in the original description of their hirsutism score. Thus, the prevalence of hirsutism in our series of women with type 1 diabetes seems to be elevated, compared not only with the prevalence in the nondiabetic population from Madrid but also with the prevalence reported elsewhere.
The high prevalence of hyperandrogenic disorders in type 1 diabetic women persisted when the analysis was restricted to the group of 68 type 1 diabetic patients who were tested for serum androgen levels, as 22.1% of these women matched criteria for the diagnosis of PCOS, and 19.1% had hirsutism without menstrual dysfunction.
This increased prevalence of hyperandrogenic disorders was accompanied by the corresponding increases in serum androgen concentrations. The group of type 1 diabetic women with PCOS showed increases in serum total and free testosterone and androstenedione concentrations. On the contrary, the group of hirsute type 1 diabetic women without menstrual dysfunction presented intermediate serum androgen concentrations that were not different from those of diabetic women with PCOS or those of nonhyperandrogenic diabetic women and healthy controls. Considering regular menses as indicative of normal ovulatory function, our present results suggest that these patients have idiopathic hirsutism (32).
We have recorded menstrual cycles but have not used more precise methods to evaluate ovulatory dysfunction. Forty percent of hirsute patients with apparently normal cycles have oligo/anovulation, when studied by basal body temperature calendars and day 22–24 progesterone levels (32). Thus, it is possible that some of the hirsute type 1 diabetic women with regular menstrual cycles in our series may have had ovulatory dysfunction, resulting in an underestimate of the true prevalence of PCOS in women with type 1 diabetes mellitus.
The normal serum DHEAS concentration in type 1 diabetic women with PCOS suggests that the ovary is the main source of androgen excess in type 1 diabetic patients. The increased ovarian androgen synthesis might result from the exogenous hyperinsulinism derived from the supraphysiological doses of insulin needed to achieve an adequate glycemic control in these patients (11). Hyperinsulinism may facilitate the LH-mediated androgen synthesis in the ovary (5) or may directly stimulate androgen secretion (4), as has been described in type 2 diabetes mellitus and other insulin-resistant states (1).
PCOS is considered as a form of functional ovarian hyperandrogenism, resulting from a dysregulation of androgen secretion (4). Virdis et al. (33) found functional ovarian hyperandrogenism, defined by an exaggerated serum 17-OHP response to the GnRH analog leuprolide, in five of nine women with type 1 diabetes mellitus and oligomenorrhea, supporting a role for the ovary as the source of androgen excess in type 1 diabetic women.
Exogenous hyperinsulinism might be favored in type 1 diabetes mellitus by the fact that it is not being delivered into the portal circulation but though a nonphysiological sc route. Because the synthesis and secretion of SHBG is suppressed by the action of insulin delivered through the portal circulation (34), the sc route of insulin delivery may also explain why SHBG levels were not decreased in women with type 1 diabetes and PCOS to the extent expected from their excessive serum androgen concentrations.
The normal SHBG levels found in hyperandrogenic type 1 diabetic women might protect these patients against androgen excess at the tissue level by determining the amount of testosterone that is available to tissues. This mechanism may provide an explanation for the mild degree of hirsutism (in only 3 of the 26 hirsute type 1 diabetic women the hirsutism score was more than 15) found in these patients.
Our data have not shown a significant influence of obesity in the development of hyperandrogenic symptoms in women with type 1 diabetes mellitus, given that the body mass index was not different among diabetic women with or without hyperandrogenic disorders.
Ideally, to explain the hyperandrogenism of type 1 diabetic women as a consequence of exogenous hyperinsulinism, the type 1 diabetic patients with PCOS or hirsutism without menstrual dysfunction should have been treated with higher insulin doses than those diabetic women without hyperandrogenic disorders. Unfortunately, retrospective analysis of clinical variables related to type 1 diabetes mellitus and its treatment has not revealed any significant predisposing factor, including those related to insulin treatment. Therefore, we cannot conclude that exogenous hyperinsulinism is the main cause of hyperandrogenism in these women. Nevertheless, exogenous hyperinsulinism might have triggered other predisposing factors for androgen excess, contributing to the markedly increased prevalence of PCOS and hirsutism in women with type 1 diabetes mellitus.
Moreover, the hypothetical influence of the clinical variables studied here on the development of hyperandrogenism might have been biased by the homogeneity of our diabetic population, because there were no marked differences in most of these parameters among the individuals described here. In future studies, comparison of grossly different diabetic populations, in terms of glycemic control or insulin dose, might disclose a possible influence of these factors on androgen excess.
In summary, women with type 1 diabetes mellitus present a marked increase in the prevalence of hyperandrogenic disorders such as PCOS and hirsutism. Although our present results need confirmation in type 1 diabetic patients from different ethnic backgrounds, and further studies are needed to define the mechanisms underlying the hyperandrogenism of these women, androgen excess should be added to the constellation of symptoms of type 1 diabetes mellitus. When these women present with reproductive dysfunction, the possibility ought be considered that the underlying cause might be, at least partly, their hyperandrogenism. In such a case, specific treatment for this condition might prove beneficial.
|
Footnotes |
|---|
Received March 1, 2000.
Revised May 15, 2000.
Accepted July 24, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  E. Codner, G. Iniguez, C. Villarroel, P. Lopez, N. Soto, T. Sir-Petermann, F. Cassorla, and R. A. Rey Hormonal Profile in Women with Polycystic Ovarian Syndrome with or without Type 1 Diabetes Mellitus J. Clin. Endocrinol. Metab., December 1, 2007; 92(12): 4742 - 4746. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Codner and H. F. Escobar-Morreale Hyperandrogenism and Polycystic Ovary Syndrome in Women with Type 1 Diabetes Mellitus J. Clin. Endocrinol. Metab., April 1, 2007; 92(4): 1209 - 1216. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Codner, N. Soto, P. Lopez, L. Trejo, A. Avila, F. C. Eyzaguirre, G. Iniguez, and F. Cassorla Diagnostic Criteria for Polycystic Ovary Syndrome and Ovarian Morphology in Women with Type 1 Diabetes Mellitus J. Clin. Endocrinol. Metab., June 1, 2006; 91(6): 2250 - 2256. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Codner, D. Mook-Kanamori, R. A. Bazaes, N. Unanue, H. Sovino, F. Ugarte, A. Avila, G. Iniguez, and F. Cassorla Ovarian Function during Puberty in Girls with Type 1 Diabetes Mellitus: Response to Leuprolide J. Clin. Endocrinol. Metab., July 1, 2005; 90(7): 3939 - 3945. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Roldan, H. F. Escobar-Morreale, R. Barrio, H. de la Calle, M. Alonso, R. Garcia-Robles, and J. Sancho Identification of the Source of Androgen Excess in Hyperandrogenic Type 1 Diabetic Patients Diabetes Care, July 1, 2001; 24(7): 1297 - 1299. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
