Androgen excess disorders remain undiagnosed in one of every four premenopausal women with Type 1 diabetes

Abstract

Study question: How frequent are androgen excess disorders, including polycystic ovary syndrome (PCOS), among women with Type 1 diabetes mellitus (T1D)?

Summary answer: One in every four women with T1D suffer from undiagnosed androgen disorders, with the classic phenotype of PCOS being the most frequent.

What is known already: Systemic iatrogenic hyperinsulinism is unavoidable in patients with T1D because insulin is administered subcutaneously instead of being secreted directly into the portal circulation. Since insulin acts as a co-gonadotrophin at the ovary, iatrogenic hyperinsulinism might trigger androgen secretion in predisposed women. Most studies conducted to date have reported increased prevalences of androgen excess disorders in premenopausal women with T1D, yet these studies were hampered by methodological limitations that preclude reaching a definite conclusion on the issue.

Study design size and duration: From January 2020 to March 2024, we conducted a cross-sectional study including women with T1D.

Participants setting methods: We recruited 149 consecutive premenopausal women with T1D who attended the diabetes clinics of an Academic Hospital at Madrid, Spain. We compared them with 295 typical patients with PCOS who did not have T1D. We used state-of-the-art mass spectrometry techniques to measure serum androgens and equilibrium dialysis to measure free testosterone and followed the latest guidelines to phenotype patients.

Main results and the role of chance: Hyperandrogenic disorders (considering PCOS, idiopathic hyperandrogenism, and idiopathic hirsutism as a whole) were present in 39 (of 149) women with T1D (26%, 95% CI: 20-34%), including 30 women who fulfilled the PCOS diagnostic criteria, indicating a prevalence of 20% (95% CI: 15-27%). The most common PCOS phenotype was the classic combination of hyperandrogenism and ovulatory dysfunction. Women with T1D and PCOS were younger (mean age 25 ± 7 vs 31 ± 9 years-old, P = 0.003) and their onset of T1D was more frequently premenarcheal (73% vs 46%, P = 0.008) compared to those without PCOS. Compared to 295 typical patients with PCOS without T1D, the 30 women with T1D and PCOS showed milder hyperandrogenic signs and lower free testosterone concentrations [13 (9, 25) vs 21 (15, 29) pM, P < 0.001] regardless of the glucose tolerance of the former.

Limitations reasons for caution: We acknowledge the possibility of selection bias: having excluded T1D women already diagnosed with PCOS, we may have underestimated actual prevalence rates. Also, the cross-sectional design of the study precluded us from obtaining any causality insights about the associations found here.

Wider implications of the findings: One in every four women with T1D suffer androgen excess disorders, with the classic combination of hyperandrogenism and ovulatory dysfunction being the most common phenotype of PCOS. Women with a premenarcheal onset of T1D are particularly susceptible to developing androgen excess disorders and may benefit from future preventive measures at young ages. Routine screening for these prevalent disorders seems reasonable to avoid the negative consequences of androgen excess and chronic ovulatory dysfunction on the general and reproductive health of T1D women.

Study funding/competing interests: This work was supported by grants PIE18/01122 and PI21/00116 from Instituto de Salud Carlos III, and co-funded by the European Union. A.B.C. is the recipient of a Río Hortega grant (CM19/00138) from Instituto de Salud Carlos III. CIBERDEM and IRYCIS also belong to Instituto de Salud Carlos III. The funding source was not involved in the study design, the data collection, analysis and interpretation, nor in the decision to submit the paper for publication. The authors have no competing interests to disclose.

Trial registration number: N/A.

Keywords: Type 1 diabetes; hirsutism; hyperandrogenism; ovulatory dysfunction; polycystic ovary syndrome.

Figure 1.

Concordance between equilibrium dialysis and calculated free testosterone (FT). Upper panel figure shows a Bland–Altman plot, with the differences (y-axis) plotted against the mean of the two measurements (x-axis). The black lines represent the limits of agreement (mean ± 1.96 SD). Bottom panel shows the linear representation of the relationship between the two assays of FT, showing with the line of identity (y = x). Lin’s Concordance Correlation Coefficient (95% CI) is shown. To convert to metric units, multiply free testosterone by 0.002884 (result in ng/dl).

Figure 2.

Box plots representing clinical, hormonal, and other laboratory parameters from the comparison between women with Type 1 diabetes and PCOS (T1D) and typical patients with PCOS presenting with abnormal glucose tolerance (AGT) or normal glucose tolerance (NGT). The median is represented by the central line in each box, while the interquartile range is captured by the box edges. Whiskers represent the range of values within 1.5 times the interquartile range, and outliers are displayed as individual points. BMI adjusted P values were determined for intergroup comparisons. Statistically significant differences in the pairwise comparisons are represented with an asterisk (*). To convert to conventional units, multiply total testosterone by 28.84 (result in ng/dl), calculated free testosterone by 0.002884 (result in ng/dl), sex hormone-binding globulin (SHBG) by 9 (result in µg/dl), dehydroepiandrosterone-sulfate (DHEAS) by 368.5 (result in ng/ml), Δ⁴-androstendione by 0.2865 (result in ng/ml), 17-OH progesterone by 0.330 (result in ng/ml), cholesterol by 38.6 (result in mg/dl), and triglycerides by 88.5 (result in mg/dl). HDL, high-density lipoproteins; LDL, low-density lipoproteins; US, ultrasensitive.