Change in Circulating Levels of Endothelial Progenitor Cells and Sexual Function in Women With Type 1 Diabetes.

CONTEXT: Endothelial progenitor cells (EPCs), which are involved in the mechanisms of vascular repair and sexual function, are decreased in diabetic women compared with general population.
OBJECTIVE: This work aimed to investigate the circulating levels of EPCs and the change in sexual function during the menstrual cycle in women with type 1 diabetes (T1DM) compared with healthy women.
METHODS: This case-control observational study was conducted at the Unit of Endocrinology and Metabolic Diseases at University Hospital "Luigi Vanvitelli'' of Naples. Participants included 36 women with T1DM and 64 age-matched healthy controls. EPCs were quantified by flow cytometry and sexual function was assessed using the Female Sexual Function Index (FSFI) and the Female Sexual Distress Scale. All assessments were made at the follicular, ovulatory, and luteal phases of the same menstrual cycle. Main outcome measures included differences in EPCs levels and sexual function between patients and controls.
RESULTS: Compared with controls, women with T1DM showed significantly lower levels of both CD34 + (P < .001) and CD34 + CD133 + cells (P < .001) in the ovulatory phase, and CD34 + KDR + cells both in the ovulatory phase and in the luteal phase (P < .001 for both). Diabetic women showed significantly lower total FSFI scores and higher FSDS score than control women in all phases of the menstrual cycle. FSFI total score was predicted by both CD34 + CD133 + and CD34 + KDR + cells in the follicular phase, CD34 + and CD34 + KDR + CD133 + cells in the ovulatory phase, and CD34 + KDR + and CD34 + KDR + CD133 + cells in the luteal phase.
CONCLUSION: Women with T1DM show lower levels of EPCs during the menstrual cycle compared with controls. EPCs count predicts sexual function in this selected population.

The global incidence of type 1 diabetes mellitus (T1DM) is increasing, with rates in children younger than 5 years being of particular concern (1). This imposes a lifelong burden and increased risk of early death from cardiovascular disease (2). Indeed, the risk of cardiovascular death is 4.2 times higher in people with T1DM compared with nondiabetic controls (3, 4).

Cardiovascular disease also disproportionally affects women with T1DM, which contrasts with a male predominance observed in the general population (5, 6). Diabetes attenuates the overall biological advantage of women by protecting them from cardiovascular complications across all ages (7). Reproductive factors, differences in experience, and presentation of symptoms or psychosocial stress may also play a role in the less favorable situation of diabetic women (7).

Hyperglycemia contributes substantially to endothelial damage, being responsible for a process of chronic inflammation, production of reactive oxygen species, and alteration of hypoxia sensors. Circulating endothelial progenitor cells (EPCs) play a key role in maintaining endothelial homeostasis because of their ability to differentiate into mature endothelial cells and participate in the mechanisms of neoangiogenesis and endothelial repair. EPCs show a wide heterogenic antigenic profile. Typical surface antigens to identify EPCs are CD34, CD133, and KDR (8). There is evidence that circulating levels of EPCs are reduced in diabetic patients compared with age-matched individuals (9, 10), representing one of the mechanisms linking increased vascular risk to diabetes mellitus (11).

During the menstrual cycle, the number of EPCs in peripheral blood undergoes a cyclic change. Robb et al (12) have demonstrated that in healthy nulliparous, premenopausal, nonsmoking women with a regular menstrual cycle, CD34 + CD133 + KDR + cells fluctuated during the menstrual cycle with midfollicular levels being 3-fold higher than periovulatory levels. Moreover, Foresta and colleagues (13) observed a significant increase in the number both of CD34 + CD133 + and CD34 + CD133 + KDR + cells in young fertile women during ovulation. The circulating levels of EPCs during the menstrual cycle in women with T1DM have never been investigated.

Sexual health is fundamental to the overall health and well-being of individuals of both sexes. A growing body of evidence suggests that erectile dysfunction (ED) is an efficient predictor of cardiovascular disease in diabetic men (14, 15). There is evidence that men with ED have a lower number of EPCs compared with healthy controls (16, 17). Of note, we previously reported that T1DM patients with ED show reduced levels of CD34 + KDR + CD133 + cells, whose number correlates with the severity of the symptom (16). On the other hand, whether sexual dysfunctions in women are associated with the risk of developing cardiovascular diseases remains unclear (18). Moreover, the relationship between circulating levels of EPCs and sexual function in women has never been investigated.

The aim of this study is to investigate the trend of circulating EPCs levels in the follicular, ovulatory, and luteal phases of the menstrual cycle and the change in sexual function in young women with T1DM during the menstrual cycle compared with age-matched, nondiabetic women.

Materials and Methods

Study Design and Participants

This is a single-center, observational study conducted between January 2019 and March 2021. Women with T1DM attending the Unit of Endocrinology and Metabolic Diseases at University Hospital “Luigi Vanvitelli’’ (Naples, Italy) were consecutively screened for eligibility criteria: 1) age 18 years and older and younger than 35 years, 2) stable couple relationship or masturbation in the previous 6 months, 3) regular menstrual cycle, and 4) absence of oral contraceptive use. We excluded patients with major health problems including diabetic chronic complications, neoplasms, major depression or other psychiatric disorders, severe neurological diseases, drug or alcohol abuse, polycystic ovarian syndrome, and use of medication with recognized adverse effects on female sexual function. Also excluded were women who were pregnant or planning to become pregnant and those who experienced gynecological surgery, lower urinary tract symptoms, and pelvic trauma in the last 6 months. All women who met the inclusion criteria and agreed to sign the informed consent were enrolled in the study. The study was approved by the local ethics committee, and all participants signed an informed consent before enrollment.

Assessment of Sexual Function

Patients and controls were asked to complete 2 validated multiple-choice questionnaires assessing sexual function (19) and the discomfort related to sexual activity (20). Scores for each instrument were calculated according to the recommended scoring system. Each questionnaire was administered after a short explanation. Sexual function was assessed by completing the Female Sexual Function Index (FSFI), a self-report questionnaire including 19 items subdivided into 6 domains (desire, arousal, lubrication, orgasm, satisfaction, and pain) referring to sexual activity in the last 4 weeks. Each domain was scored on a scale of 0 or 1 to 6, with a higher score indicating better function. For each of the 6 domains, a score was calculated and the total score was obtained by adding the scores of 6 domains. The total score ranged from 2 to 36, and impaired sexual function was indicated by a score of 26.55 or less. Sexual activity–related distress was assessed using the Female Sexual Distress Scale (FSDS), a self-assessment questionnaire composed of 12 items. Women were required to quantify the frequency of each domain score on a scale of 0 to 4 (0 = never, 4 = always). A total score higher than or equal to 15 indicated distress related to sexual life. In accordance with the American Psychiatric Association guidelines, each sexual function domain was considered altered if associated with personal distress. Female sexual dysfunction (FSD) was diagnosed according to a FSFI score lower than 26.55 and an FSDS score higher than 15 (21).

Clinical Measures and Laboratory Analyses

All patients underwent a full physical examination to assess weight and height, body mass index (BMI), and blood pressure. Height and weight were measured to the nearest 0.5 cm and 100 g, respectively, with participants wearing lightweight clothing and no shoes. BMI was calculated as weight (in kilograms) divided by standing height (in meters squared). Arterial blood pressure was measured 3 times at the end of the physical examination with the participant in the sitting position. Assays for fasting glucose, glycated hemoglobin A1c (HbA1c), total cholesterol, low-density and high-density lipoprotein cholesterol, and triglyceride levels were performed in the hospital’s chemistry laboratory. Blood samples were drawn in the follicular, ovulatory, and luteal phases of the same menstrual cycle to assess sex hormone levels, including follicle-stimulating hormone (FSH), luteinizing hormone (LH), progesterone, and estradiol.

Assessment of Circulating Levels of Endothelial Progenitor Cells

Peripheral blood cells were analyzed for the expression of surface antigens by direct flow cytometry, as previously described (22). Briefly, fasting blood samples were processed after 1 to 2 hours. Mononuclear cells were isolated from peripheral venous blood by density centrifugation. Then, the isolated blood cells were stained for 30 minutes at 4 °C in the dark with fluorescein isothiocyanate (FITC)-conjugated antihuman CD34 monoclonal antibody (mAb) (BD Biosciences catalog No. 555821, RRID:AB_396150, https://scicrunch.org/resolver/RRID:AB_396150), phycoerythrin (PE)-conjugated antihuman KDR mAb (R&D Systems catalog No. FAB357P, RRID:AB_357165, https://scicrunch.org/resolver/RRID:AB_357165), and allophycocyanin (APC)-conjugated antihuman CD133 (Miltenyi Biotec catalog No. 130-090-664, RRID:AB_244341, https://scicrunch.org/resolver/RRID:AB_244341). Isotope immunoglobulin IgG1 and IgG2a antibody was used to discriminate between signal range and baseline fluorescence within the samples. After incubation, quantitative analysis was performed on a BD FACSCalibur cytometer, and 1 000 000 cells were acquired in each sample. A morphological gate was used to exclude granulocytes. Then, we gated CD34+ or CD133+ peripheral blood cells in the mononuclear cell fraction and examined the resulting population for the dual expression of KDR. In the 2-dimensional dot-plot analysis, we identified CD34+ CD133+ cells. Triple-positive cells were identified by the dual expression of KDR and CD133 in the CD34+ gate. Data were processed with the use of the Macintosh CELLQuest software program (Becton Dickinson). The results from flow cytometry were expressed as the number of cells per 106 events.

Statistical Analysis

Data in the tables are presented as mean ± SD, median, and interquartile range (IQR), or number and percentage. The Kolmogorov-Smirnov test was used to assess if variables were normally distributed. Descriptive statistics were used for demographic and baseline clinical characteristics of all participants in the study. Comparisons of baseline data between the patient groups were performed by t test or Mann-Whitney rank sum test, depending on the normality of sample distribution. The χ2 test was used for comparing dichotomous variables. The correlation between EPCs cell count and clinical variables was assessed using the Spearman coefficient of correlation. Multiple regression analysis was conducted to characterize the association between circulating levels of the 4 evaluated EPCs phenotypes (CD34+, CD34+ CD133+, CD34+ KDR+, CD34+ KDR+ CD133+ cell count) and the FSFI total score. Data were analyzed using Stata, version 16.0 (Stata Corp). A P value less than .05 was considered statistically significant.

Results

A total of 112 individuals were screened for eligibility and 4 women were excluded (1 for polycystic ovarian syndrome, and 3 for estroprogestinic use). Eight women did not give consent to participate in the study. Therefore, a total of 36 women with T1DM and 64 healthy controls were included in the study. The clinical and metabolic characteristics of the entire study population are shown in Table 1. The mean age was 25 years and the mean BMI was 24.3. The mean age did not differ between women with T1DM and control women, nor did the anthropometric parameters, lipid profile, or blood pressure. As expected, women with T1DM had higher fasting glucose (P < .001) and HbA1c levels (P < .001), compared with the control group. Diabetic women were treated with multiple daily insulin injections (n = 12) or continuous subcutaneous insulin infusion (n = 24) combined with continuous glucose monitoring. The overall prevalence of FSD, defined as the simultaneous presence of pathological FSFI and FSDS scores, was 11% (4/36) among women with T1DM and 12.5% (8/64) among control women.

Table 1.

Characteristics of participants in the study

Parameters	Women with type 1 diabetes (N = 36)	Controls (N = 64)	P
Age, y	25.0 ± 5.6	24.8 ± 2.3	.805
Duration of diabetes, y	14.1 ± 4.8	–	–
MDI/CSII, n	12/24	–
Weight, kg	62.9 ± 15.2	60.7 ± 7.3	.336
BMI	25.8 ± 5.0	23.7 ± 2.6	.120
Fasting glucose, mg/dL	133.6 ± 42.9	76.0 ± 5.6	< .001
HbA1c, %	7.7 ± 1.1	5.0 ± 0.6	< .001
SBP, mm Hg	110.0 (110.0-112.5)	110.0 (105.0-115.0)	.790
DBP, mm Hg	70.0 (65.0-70.0)	67.5 (62.5-70)	.331
Total cholesterol, mg/dL	168.5 ± 20.8	163.0 ± 20.4	.193
HDL cholesterol, mg/dL	60.7 ± 13.3	58.1 ± 8.4	.233
LDL cholesterol, mg/dL	96.3 ± 15.7	101.2 ± 26.0	.288
Triglycerides, mg/dL	57 (46, 65)	61 (51, 66)	.098
Creatinine, mg/dL	0.80 ± 0.1	0.81 ± 0.01	.443

Data are expressed as mean ± SD and as median (interquartile range).

Abbreviations: BMI, body mass index; CSII, continuous insulin infusion; DBP, diastolic blood pressure; HbA1c, glycated hemoglobin A1c; HDL, high-density lipoprotein; LDL, low-density lipoprotein; MDI, multiple daily injections; SBP, systolic blood pressure.

In both groups, there was a normal cyclical variation in circulating pituitary and ovarian hormones with periovulatory peaks both in serum LH and FSH concentrations, and periovulatory and midluteal peaks both in serum estradiol and progesterone concentrations. There were no differences in sex hormone levels between the 2 groups during the different phases of the menstrual cycle (Table 2).

Table 2.

Sex hormone levels during follicular, ovulatory, and luteal phases of the menstrual cycle in women with type 1 diabetes and healthy controls

	Follicular phase			Ovulatory phase			Luteal phase
	Type 1 diabetic women	Control group	P	Type 1 diabetic women	Control group	P	Type 1 diabetic women	Control group	P
FSH, mIU/mL	7.1 ± 1.8	6.8 ± 1.6	.385	6.05 ± 2.2	6.1 ± 2.6	.921	3.6 ± 1.5	4.0 ± 1.8	.248
LH, mIU/mL	5.3 ± 1.5	5.7 ± 1.6	.212	8.7 ± 4.1	10.5 ± 6.1	.153	4.0 ± 1.5	3.7 ± 1.4	.310
Progesterone, ng/mL	0.42 ± 0.2	0.36 ± 0.17	.111	2.2 ± 3.9	1.7 ± 0.9	.341	6.0 ± 4.8	4.8 ± 2.4	.118
Estradiol, pg/mL	65.0 ± 61.1	77.4 ± 70.3	.365	133.2 ± 72.7	129.6 ± 74.3	.811	137.1 ± 77.9	140.0 ± 71.6	.848

Abbreviations: FSH, follicle-stimulating hormone; LH, luteinizing hormone. Data are axpressed as mean ± SD.

Table 3 shows the change in circulating levels of EPCs in women with T2DM and control women during the 3 phases of the menstrual cycle. Compared with controls, women with T1DM showed significantly lower levels of CD34+ cells in the ovulatory phase (cases vs control, median and IQR, 211 [171-310] vs 361 [345-437]; P < .001) (Fig. 1A), and CD34+ KDR+ cells both in the ovulatory phase (22 [19-26] vs 41 [17-251]; P < .001) and in the luteal phase (19 [9-27] vs 38 [28-49]; P < .001) (Fig. 1B). In the ovulatory phase, the CD34+ CD133+ cell count was significantly lower in women with diabetes as compared with controls (111 [80-113] vs 153 [110-218]; P < .001) (Fig. 1C). No significant difference was observed for the other EPCs phenotypes between the 2 groups (see Table 3).

Table 3.

Circulating levels of endothelial progenitor cells in the follicular, ovulatory, and luteal phases in women with type 1 diabetes and healthy controls

	Women with type 1 diabetes (N = 36)	Control group (N = 64)	P
CD34+
Follicular phase	207 (171-320)	217 (147-330)	.646
Ovulatory phase	211 (171-310)	361 (345-437)	< .001
Luteal phase	247 (166-267)	219 (142-296)	.343
CD34 + CD133+
Follicular phase	119 (90-132)	124 (76-198)	.626
Ovulatory phase	111 (80-113)	153 (110-218)	< .001
Luteal phase	114 (58-14)	116 (15-7)	.343
CD34 + KDR+
Follicular phase	23 (12-32)	28 (23-47)	.082
Ovulatory phase	22 (19-26)	41 (29-251)	< .001
Luteal phase	26 (9-27)	38 (28-49)	< .001
CD34 + KDR + CD133+
Follicular phase	7 (2-10)	8 (6-11)	.087
Ovulatory phase	4 (3-6)	5 (4-7)	.133
Luteal phase	6 (2-7)	7 (3-12)	.120

Data are expressed as median (interquartile range).

Figure 1.

Circulating levels of A, CD34+; B, CD34+ KDR+; and C, CD34+ CD133+ cells in the 3 phases of the menstrual cycle in women with diabetes and healthy controls. *P significant vs controls.

Patients with T1DM showed significantly lower total FSFI scores and higher FSDS scores than control women in all 3 phases of the menstrual cycle (Table 4). Compared with control women, diabetic women had significantly lower scores in the domains of desire, arousal, and pain in the follicular phase, and significantly lower scores in all the domains both in the ovulatory and luteal phases (see Table 4).

Table 4.

Single domains scores during the different phases of the menstrual cycle in diabetic women and healthy controls

	Women with type 1 diabetes (N = 36)	Control group (N = 64)	P
Desire
Follicular phase	4.2 (3.6-4.8)	5.1 (4.5-5.7)	< .001
Ovulatory phase	4.2 (3.6-4.2)	5.4 (4.8-6)	< .001
Luteal phase	4.8 (3.6-4.8)	5.4 (4.8-6)	< .001
Arousal
Follicular phase	5.1 (4.2-5.4)	5.7 (5.5-5.8)	< .001
Ovulatory phase	5.1 (4.2-5.4)	6 (5.7-6)	< .001
Luteal phase	4.5 (4.2-5.4)	6 (6-6)	< .001
Lubrication
Follicular phase	4.8 (3.6-6)	5.7 (5.1-6)	.002
Ovulatory phase	5.4 (4.2-5.4)	6 (5.7-6)	< .001
Luteal phase	4.5 (3.6-5.4)	6 (6-6)	< .001
Orgasm
Follicular phase	4.4 (3.2-5.6)	4.8 (4.4-5.4)	.098
Ovulatory phase	4.8 (3.2-5.2)	5.6 (5.6-6)	< .001
Luteal phase	4.8 (4-5.6)	6.6 (6-6)	< .001
Pain
Follicular phase	5.6 (4-6)	6 (6-6)	< .001
Ovulatory phase	6 (4.8-6)	6 (6-6)	< .001
Luteal phase	5.6 (4-6)	6 (6-6)	< .001
Satisfaction
Follicular phase	5.2 (4-6)	5.4 (4.8-6)	.159
Ovulatory phase	5.6 (4-6)	6 (6-6)	.001
Luteal phase	5.2 (4.4-5.6)	6 (6-6)	< .001
FSFI total score
Follicular phase	28.4 (22-32.8)	32.6 (31.5-33.3)	< .001
Ovulatory phase	28.4 (26.8-31.9)	34.8 (34.6-35.3)	< .001
Luteal phase	29.4 (25-29.9)	35.4 (34.8-36)	< .001
FSDS total score
Follicular phase	6 (2-8)	1 (0-3)	<0.001
Ovulatory phase	3 (1-9)	1 (0-1)	<0.001
Luteal phase	5 (4-11)	1 (0-1)	<0.001

Data are expressed as median (interquartile range).

Abbreviations: FSDS, Female Sexual Distress Scale; FSFI, Female Sexual Function Index.

Results from the univariate analysis are shown in Tables 5 to 7. In the overall population, FSFI total score was positively associated both with FSH and estradiol levels in all 3 phases of the menstrual cycle (see Table 5); no significant association were found between sex hormone levels and FSDS score (see Table 5). Moreover, a significantly negative correlation was found between HbA1c and FSH and LH levels, FSFI total score, and CD34+, CD34+ KDR+, and CD34+ KDR+ CD133+ count in all 3 phases of the menstrual cycle (see Table 6).

Table 5.

Correlation between sex hormones and Female Sexual Function Index and Female Sexual Distress Scale score in the different phases of the menstrual cycle in the overall population. Numbers in bold indicate statistical significant correlations

FSFI	Follicular phase		Ovulatory phase		Luteal phase
	rSp	P	rSp	P	rSp	P
FSH	0.340	< .001	0.349	.003	0.312	.002
LH	0.062	.557	0.038	.710	–0.018	.857
Estradiol	0.382	.005	0.369	.006	0.280	.005
Progesterone	–0.070	.289	0.173	.102	0.003	.898
FSDS
FSH	0.040	.765	0.098	.493	0.021	.712
LH	0.054	.568	0.066	.520	–0.039	.768
Estradiol	–0.192	.538	–0.098	.832	–0.036	.632
Progesterone	–0.046	.372	0.149	.236	0.006	.841

Abbreviations: FSDS, Female Sexual Distress Scale; FSFI, Female Sexual Function Index; FSH, follicle-stimulating hormone; LH, luteinizing hormone.

Table 6.

Correlation between glycated hemoglobin A1c, sexual function indices, and endothelial progenitor cell total count in the different phases of the menstrual cycle in the overall population. Numbers in bold indicate statistical significant correlations

HbA1c	Follicular phase		Ovulatory phase		Luteal phase
	rSp	P	rSp	P	rSp	P
FSH	–0.234	.021	–0.403	< .001	–0.409	< .001
LH	0.162	.123	–110	.289	–0.189	.064
Estradiol	–0.326	.001	–0.201	.048	–0.461	< .001
Progesterone	0.029	0.378	0.173	.102	0.032	.808
FSFI	–0.637	< .001	–0.468	< .001	–0.588	< .001
FSDS	–0.047	.642	–0.099	.345	–0.071	.483
CD34+	–0.229	.024	–0.477	< .001	–0.273	.007
CD34 + 133+	0.156	.127	–0.286	.011	0.047	.645
CD34 + KDR+	–0.342	< .001	–0.455	< .001	–0.477	< .001
CD34 + KDR + 133+	–0.416	< .001	–0.401	< .001	–0.536	< .001

Abbreviations: FSDS, Female Sexual Distress Scale; FSFI, Female Sexual Function Index; FSH, follicle-stimulating hormone; HbA1c, glycated hemoglobin A1c; LH, luteinizing hormone.

In women with T1DM, circulating levels of CD34+ cells were negatively associated with FSFI total score in the follicular phase and in the luteal phase. (r = –0.483, P = .003; r = –0.367, P = .028). CD34+ KDR+ cell count was inversely correlated with FSFI total score in the luteal phase (r = –0.400, P = .015) (see Table 7). Moreover, circulating levels of CD34+ CD133+, and CD34+ KDR+ CD133+ cells were negatively associated with FSFI total score in the ovulatory phase (r = –0.703, P < .001) and follicular phase (r = –0.452, P = .006), respectively (see Table 7).

Table 7.

Correlation between circulating levels of CD34+, CD34 + KDR+, CD34 + CD133 + and CD34 + CD133 + KDR + cells and Female Sexual Function Index total score in the different phases of the menstrual cycle in women with diabetes and in healthy controls. Numbers in bold indicate statistical significant correlations

Women with type 1 diabetes (N = 36)
FSFI total score	Follicular phase		Ovulatory phase		Luteal phase
	rSp	P	rSp	P	rSp	P
CD34+	–0.483	.003	–0.075	.660	–0.367	.028
CD34 + KDR+	–0.050	.771	0.092	.590	–0.400	.015
CD34 + CD133+	–0.056	.732	–0.703	< .001	0.08	.626
CD34 + CD133 + KDR+	–0.452	.006	–0.084	.624	–0.157	.378
Control women (N = 64)
CD34+	0.600	< .001	0.702	< .001	0.547	< .001
CD34 + KDR+	0.672	< .001	–0.034	.791	0.738	< .001
CD34 + CD133+	0.596	< .001	–0.095	.536	0.730	< .001
CD34 + CD133 + KDR+	–0.214	.742	0.750	< .001	0.034	.245

Abbreviation: FSFI, Female Sexual Function Index.

In the control group, the number of CD34 + cells was associated with total FSFI score in the follicular phase (r = 0.600, P < .001), in the ovulatory phase (r = 0.702, P < .001), and in luteal phase (r = 0.547, P < .001) (see Table 7). CD34+ KDR+ cell count was associated with total FSFI score in the follicular phase (r = 0.672, P < .001) and in the luteal phase (r = 0.738, P < .001) (see Table 7). Moreover, circulating levels of CD34+ CD133+ cells were associated with total FSFI score in the follicular phase (r = 0.596, P < .001) and in the luteal phase (r = 0.730, P < .001) (see Table 7). In the multivariable regression analysis, total FSFI score was predicted both by CD34+ CD133+ (β coefficient 1.136, P < .001) and CD34+ KDR+ (β coefficient 0.288, P = .011) cells in the follicular phase, CD34+ (β coefficient 0.342, P < .001), CD34+ KDR+ CD133+ (β coefficient 0.227, P = .045) cells in the ovulatory phase, and CD34+ KDR+ (β coefficient 0.438, P < .001) and CD34+ KDR+ CD133+ (β coefficient 0.275, P = .029) cells in the luteal phase (Table 8).

Table 8.

Statistical associations between Female Sexual Function Index total score and endothelial progenitor cell count by multiple linear regression. Numbers in bold indicate statistical significant associations

	β Coefficient	P
Follicular phase
CD34+	–1.367	.567
CD34 + CD133+	1.136	< .001
CD34 + KDR+	0.288	.011
CD34 + KDR + CD133+	–0.009	.922
Ovulatory phase
CD34+	0.342	.011
CD34 + CD133+	–0.297	.241
CD34 + KDR+	0.175	.129
CD34 + KDR + CD133+	0.227	.045
Luteal phase
CD34+	–0.133	.365
CD34 + CD133+	0.193	.062
CD34 + KDR+	0.438	< .001
CD34 + KDR + CD133+	0.275	.029

Discussion

We have demonstrated for the first time a decrease in circulating levels of EPCs and worse sexual function in young fertile women with T1DM during the menstrual cycle. In addition, total FSFI score was predicted by EPCs cell count in the 3 phases of the menstrual cycle, suggesting the existence of a link between mechanisms of vascular repair and sexual function in young premenopausal women. Our results are novel, as no previous study has examined the change in circulating progenitor cells and its relationship with sexual function in this selected population at all phases of the menstrual cycle.

Women of reproductive age are exposed to a lower cardiovascular risk than men of the same age (5, 6). This is generally attributed to the differences in sex hormones and, specifically, to the protective cardiovascular properties of female estrogens (7). In addition to the effects on plasma lipids and the vessel wall, other mechanisms may link sexual hormones to a favorable cardiovascular profile, including the regulation of endothelial homeostasis. Interestingly, both quantitative and qualitative differences in EPCs between young men and women, together with the observation that EPCs are mobilized during the hormonal cycle, indicate that EPCs may be influenced by female sex hormones (23, 24). Indeed, there is evidence that the menstrual cycle in healthy, normally menstruating women plays a physiological role in regulating the availability of EPCs at different stages (23, 25). Our results confirm these findings; however, we found lower CD34+ and CD34+ CD133+ levels in the ovulatory phase in diabetic women compared with the control group, associated with reduced CD34+ KDR+ levels both in the ovulatory and luteal phases, suggesting that the cyclic mobilization of EPCs during the menstrual cycle observed in healthy women is attenuated in diabetic women, even in the context of a normal cyclic variation of circulating pituitary and ovarian hormones. This is relevant because it may represent one potential mechanism explaining the greater excess risk of vascular events that burden women with T1DM, compared with men (25). The main clinical determinants for such excess risk are uncertain; however, studies suggest that the deterioration of glucose control may reverse the more favorable cardiovascular risk profile of women (26). In addition, hyperglycemia could interfere with the activity of estrogen receptors and inhibit any potential protective effects on the vascular wall, promoting oxidative stress, vasoconstriction, and platelet activation (26).

A potential hypothesis explaining our findings may be the loss of the regulatory function of sex hormones on the fluctuations of EPCs levels during the phases of the menstrual cycle in women with diabetes. In type 2 diabetic women, an imbalance between estrogen receptor α/β distribution has been described that is responsible for increased vasoconstriction and enhanced vascular inflammation (27). Moreover, in postmenopausal healthy women, with the cessation of ovarian and endometrial activity, the number of EPCs are similar to those expressed in men (28). There is currently no evidence for women with T1DM. However, recent studies have shown that circulating levels of EPCs are lower in young people with T1DM compared with healthy age-matched controls (22, 23). Furthermore, in a population of young adults with T1DM, we previously identified sex differences in the count of circulating EPCs favoring males (29), suggesting that additional adverse conditions may be present in women with diabetes that further attenuate the physiological mobilization of these stem cells from the bone marrow. Whether the sex disparity in EPCs number may reflect a vulnerability for diabetic women in terms of increased risk of vascular complications remains unknown.

Our data confirm the observation that T1DM affects several aspects of female sexual function, including desire, arousal, lubrication, and pain. Indeed, we found a statistically significant reduction in the score of all domains in diabetic women compared with controls both in the ovulatory and luteal phases of the menstrual cycle. The total FSFI score as well as the scores related to desire, arousal, lubrication, and pain were lower in diabetic women compared with those of control individuals in the 3 phases of the menstrual cycle, even though they cannot be considered severely compromised. Interestingly, we reported for the first time the existence of an association between total FSFI score and the circulating levels of certain phenotypes of EPCs, highlighting a potential relationship between sexual function and angiogenesis in women. Of note, the inverse association between the count of certain EPCs phenotypes and total FSFI score in women with T1DM may suggest that the relationship that links endothelial homeostasis to sexual function is totally reversed by diabetes in women. It should be hypothesized that, in women with T1DM, the change in circulating levels of EPCs may express the effort to compensate the reduced sexual function of the diabetic women. However, even in view of the mild reduction of FSFI scores of women with T1DM and the low rate of FSD, this relationship should be worthy of investigation in mechanistic studies. Whether sexual function may be considered a reflection of the global cardiovascular health in women remains to be clarified.

Major strengths of this study include the use of validated tools for the evaluation of sexual dysfunction, and the assessment of 4 different EPCs phenotypes by flow cytometry in different phases of the same menstrual cycle. This study also has limitations. Owing to the observational nature of this study, we cannot draw conclusions about cause and effect. Moreover, a major limitation of the study relates to the limited number of individuals investigated, which needs to be extended.

In conclusion, women with T1DM show lower levels of EPCs during the ovulatory and luteal phases of the menstrual cycle, which are associated with poorer sexual function, compared with healthy controls. EPCs counts predicted total FSFI score, suggesting for the first time a relationship between angiogenesis and sexual function in young premenopausal women. Further longitudinal studies are needed to clarify whether these dysfunctions may be related to the increased cardiovascular risk of women with T1DM.