In our study we found a high prevalence of PCOS in young women with type 1 diabetes and this was higher than that reported in the general Indian population (3.7–22.5%) [32]. In contrast, another study from India [33] did not find a single case of PCOS in 58 women with T1DM, though diagnosis of PCOS was based on clinical features alone without any laboratory studies so they might have missed PCOS diagnosis in some. Moreover, a third of their study population was below 13 years of age.
High prevalence of menstrual irregularities, hyperandrogenism and PCOS has been reported in females with T1DM in other populations too [4], [5], [6], [7], [8], [9], [10], [11], [12], [13]. Results of our study are consistent with a large meta-analysis [34] including 478 girls and women (in which 24% of them had PCOS comparable to 27% in current study). Prevalence of androgen excess (25% and 26%), menstrual dysfunction (24% and 21% respectively) was also comparable (in the meta-analysis and current study. In our study fewer patients had hirsutism and higher number of subjects had PCOM (25% and 3%, 33% and 52% in meta-analysis Vs current study respectively) which may be due to racial difference, and observer and instrument differences respectively.
We did not find any difference in T1DM females with PCOS and those without PCOS with respect to parameters like median age of diabetes onset, age of menarche, HbA1c, total and unit/kg insulin requirements, BMI, duration of diabetes etc. Similar results were seen in earlier studies too [4], [9].
Unlike our study, analysis from a large prospective registry [10] of women with T1DM and PCOS (compared to T1DM alone) had significantly higher age of diabetes onset (14.9 ± 8.2 vs 11.8 ± 7.0 years p < 0.001), and higher BMI Z- score (0.92 ± 0.11 Vs 0.38 ± 0.01, p < 0.001).
One study in the Chilean population [5] found higher prevalence of hyperandrogenism, hirsutism, PCOM and PCOS than seen in our study. The proportion of women using intensive insulin therapy was higher in the PCOS/PCOM group, but there was no difference in the mean daily insulin dose in the two groups. Additionally, testosterone and androstenedione levels were also not different in these treatment groups. So, cause of this difference remained elusive.
In our study we did not find any difference in females with onset of T1DM before and after menarche. It has been hypothesized that premenarchal onset of T1DM may reprogram ovaries to produce more androgen due to high insulin exposure. It was suggested that in the gonadotropin independent phase (prepubertal), insulin enhances noncyclic recruitment of early stages of follicles which may lead to increased androgen production in later life [36], [37], but like current study Amato et al. [9] too did not find any difference in proportion of PCOS in pre or post pubertal onset of T1DM. Moreover, a recent Polish study [35] examining this specific question did not show any difference in prevalence of PCOS, hormonal profile and ovarian parameters in women with onset of T1DM before or after menarche.
When females with T1DM and PCOS were compared with females with PCOS only, prevalence of hirsutism was less in T1DM with PCOS compared to PCOS only group (3% vs 20%) with lower FG scores. Other researchers too have observed lower prevalence of hirsutism in PCOS with T1DM population compared to PCOS only women [8], [38] and can be explained by high SHBG levels leading to low free testosterone levels.
Our study showed a statistically significant higher level of SHBG in T1DM and PCOS group compared to PCOS only group (Median, IQR; 54.4, 38-86.2 vs 28.3, 20.4–37.4 nmol, p-0.004), similar to other studies. This finding is in concordance with the proposed mechanism of increased prevalence of hyperandrogenemia and PCOS in T1DM females. Subcutaneous administration of insulin maintains adequate levels of insulin in portal circulation to suppress glucose output but leads to higher insulin in systemic circulation leading to higher androgen production from ovaries and possibly from adrenal glands. This is in contrast to general PCOS women where higher insulin levels in portal circulation causes suppression of SHBG synthesis leading to higher levels of available free androgen as main mechanism behind hyperandrogenemia. We were unable to ascertain cause of the high prevalence of PCOS in our cohort as it was not related to age of onset of diabetes, pre or post pubertal onset of diabetes, metabolic control or the daily insulin dose. So, association of hyperandrogenemia and PCOS in T1DM has been reiterated in current study, the hypothesis that hyperinsulinemia could cause hyperandrogenism could not be explained.
The limitations of our study were small number of study subjects and androgens were measured by radio immunoassay (not by the gold standard LC-MS/ MS method). Transabdominal ultrasound was performed as majority of our subjects were unmarried and use of transvaginal ultrasound was deemed inappropriate.
Nevertheless, this is the first study from ethnically and geographically different Indian subcontinent estimating prevalence of PCOS among T1DM population and showing clinical, biochemical and sonographic profile of these females compared to age and BMI matched controls.