• Partial epilepsy;
  • Estradiol;
  • Progesterone;
  • Sex hormone binding globulin;
  • Antiepileptic drugs;
  • Seizure frequency


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Purpose: Neuroactive sex steroids influence neuron excitability, which is enhanced by estradiol (E2) and decreased by progesterone (Pg). In epilepsy, the production, metabolism, biologic availability, and activity of sex hormones may be affected by seizures themselves or by antiepileptic drugs (AEDs). This cross-sectional observational study was aimed at evaluating the relationships between sex steroids, seizure frequency, and other clinical parameters in women with partial epilepsy (PE) on AED treatments.

Methods: Serum E2, Pg, sex hormone binding globulin (SHBG) levels, free E2 (fE2), and E2/Pg ratios were determined during the follicular and luteal phases in 72 adult women with PE, and in 30 healthy controls. Hormonal data were correlated with seizure frequency, age, body weight, body mass index (BMI), disease onset and duration, and AED therapies.

Results: In patients, E2, fE2, and Pg levels were lower in both ovarian phases, whereas those of SHBG were higher than in controls. No significant changes in hormone levels and in prevalence of anovulatory cycles were observed between patients grouped according to their seizure frequency. However, when compared with those in healthy controls, luteal fE2 and Pg levels were chiefly impaired in women with more frequent seizures, mostly undergoing AED polytherapies, but not in those with absent or rarer seizures.

Conclusions: The actual changes in sex steroid levels and E2/Pg ratios did not explain an increased seizure frequency in adult women with AED-treated PE, but patients with more severe disease showed more relevant changes in their sex hormone profile and impaired Pg levels during the luteal phase.

Sex hormones cross the blood–brain barrier and act as “neuroactive steroids” on the brain (Belelli et al., 2006). In this setting, estrogens enhance neuron excitability, whereas progesterone (Pg) and other progestins—allopregnanolone in the first instance—stabilize membrane depolarization and exert anticonvulsant effects (Bäckström, 1976; Beyenburg et al., 2001; McEwen, 2002; Maguire et al., 2005; Scharfman & MacLusky, 2006).

Therefore, it has been hypothesized that spontaneous fluctuations in sex hormone levels and reproductive endocrine disorders, which are unusually common both in treated and untreated epilepsy, may conversely influence seizure occurrence (Herzog, 2008a,b). We have recently found that sex steroid levels, although globally decreased, did not correlate with seizure frequency in consecutive women with different types of epileptic syndrome, hormone changes being rather explained by antiepileptic drug (AED) treatments (Galimberti et al., 2009).

To avoid confounding factors related to different epilepsy types, in this cross-sectional observational study we have analyzed serum sex steroid levels and the ratios between serum estradiol (E2) and Pg levels throughout the ovarian cycle in women with partial epilepsy (PE), with an aim to assess: (1) differences in sex hormone levels versus healthy controls; and (2) the possible relationships between sex steroid set up, seizure frequency, and other clinical parameters. In fact, the elucidation of the role of hormone changes in influencing seizures and health in women with epilepsy might make possible an improved comprehensive management and treatment of these patients.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References


Seventy-two consecutive outpatients, age 16–47 years, attending the Epilepsy Centre, IRCCS ‘C. Mondino Institute of Neurology’ (Pavia – Italy), affected by cryptogenic or remote symptomatic PE according to the International League Against Epilepsy (ILAE) classifications (1989), participated in this study. Criteria of exclusion were: endocrine, metabolic, or neurologic diseases other than PE; previous ovariectomy; ongoing or more recent than one year’s pregnancy or lactation; estroprogestin or other hormonal therapies in the previous 6 months; and administration of drugs interfering with hypothalamic–pituitary function and/or with steroid synthesis and metabolism, except for AEDs.

Epilepsy onset had occurred between 1 and 37 years of age and lasted from 6 months to 37 years. Seizure occurrence over the previous 6 months before the study was recorded by diary schedules, which included charting of menstrual cycles. Seizure frequency was codified for the analysis by the following four-point seizure frequency score (SFS): 1, absent; 2, sporadic (one seizure per month or less); 3, frequent (four seizures per month or less); 4, very frequent (more than one seizure per week up to one or more seizure daily) (Galimberti et al., 2005, 2009). All women menstruated and ovary cycle lengths ranged between 25 and 34 days.

Forty-five subjects were on AED monotherapy: 41 were taking enzyme-inducing AEDs, that is, phenobarbital, PB (n = 10); carbamazepine, CBZ (n = 26); phenytoin, PHT (n = 2); primidone, PRM (n = 1); and oxcarbazepine, OXC (n = 2); and four non-enzyme-inducing AEDs, that is, valproic acid, VPA (n = 1) and lamotrigine, LTG (n = 3). Twenty-seven patients were on AED polytherapy: 22 with more than one enzyme-inducing AED, that is, CBZ + PB (n = 14), PB + PHT (n = 3), CBZ + PHT (n = 3), CBZ + PB + PHT (n = 1), and PRM + OXC (n = 1). Five cases were treated with both enzyme- and non-enzyme-inducing AEDs (two with PB + VPA, and single cases with PHT + CBZ, PB + CBZ + LTG, and PB + CBZ + VPA).

Ovulation and luteal function were assessed by a single Pg determination in the mid-luteal phase. A surge in steroid levels higher than 15.9 nmol/L (5.0 ng/ml) was thought to be indicative of ovulation. Anovulation was diagnosed when Pg levels were below 6.36 nmol/L (2.0 ng/ml), whereas the term of luteal phase defect was adopted for Pg levels ranging between 6.37 and 15.8 nmol/L.

Data from patients were compared with those from 30 healthy drug-free women, recruited from the hospital staff, age 19–47 years, with normal menstrual cycles (26 to 28-day length). Clinical and demographic data in controls and patients taken as a whole and divided according to their SFS are shown in Tables 1 and 2. The study was approved by the local ethics committee and written informed consent was obtained from all participants.

Table 1.   Clinical, biometric, and luteal progesterone (Pg) data in women with partial epilepsy (PE), and in healthy controls
 PE patients (n = 72)Healthy women (n = 30)
  1. Data are shown as mean values and 95% CI (confidence interval), or as number of cases and percent prevalence.

  2. AED, antiepileptic drug; BMI, body mass index; SFS, seizure frequency score.

Age (years)30.2 28.3–32.228.3 24.7–30.9
Body weight (kg)60.0 57.2–62.760.5 58.4–62.6
BMI (kg/m2)23.4 22.4–24.323.0 22.3–23.6
Disease onset (years)15.0 13.1–17.0 –
Disease duration (years)15.4 13.3–17.5 –
Mean SFS2.2 1.9–2.5 –
AED polytherapy27 37.5% –
Pg < 6.36 nmol/L17 23.6% 0  0%
Pg 6.37–15.8 nmol/L11 15.3% 0  0%
Table 2.   Clinical, demographic, and biometric data in 72 women with partial epilepsy divided according to their seizure frequency scores (SFSs)
 SFS 1SFS 2SFS 3SFS 4p-value
  1. Data are shown as mean values and 95% CI (confidence interval), median and range for seizure number over a 6-month period, or percent prevalence. Age ranges in subgroups are shown within brackets. Statistical analysis by analysis of variance (ANOVA) test or Fisher’s exact test for the analysis of distribution. Letters in apex show significant differences (p < 0.05) between SFS (seizure frequency score) groups analyzed by the Bonferroni-Dunn’s post hoc test.

  2. AED, antiepileptic drug; BMI, body mass index; Pg, progesterone.

No. of cases261427 5 
Age (years)29.5 25.9–33.1 (16–45)28.2 24.8–31.6 (20–38)32.7 29.3–36.1 (20–47)26.2 20.0–32.4 (21–32)NS
Disease onset (years)18.6a,b 15.4–21.815.6 11.6–19.712.6a  9.3–15.9 7.8b  3.0–12.60.0092
Disease duration (years)10.9a  7.4–14.412.9b 10.0–15.720.6a,b 17.1–24.018.4  9.6–27.20.0003
Seizure number/6 month period 0 – 2  1–612  7–2336 30–124
Body weight (kg)59.2 53.9–64.460.4 54.0–66.859.7 55.4–63.964.4 46.3–82.5NS
BMI (kg/m2)23.2 21.5–25.023.8 21.0–26.623.2 21.6–24.723.7 17.6–29.8NS
AED polytherapy 6 23.1% 5 35.7%12 44.4% 4 80.0%0.0532
Luteal Pg < 6.36 nmol/L 5 19.2% 4 28.6% 7 25.9% 1 20.0%NS
Luteal Pg 6.37–15.8 nmol/L 3 11.5% 1 7.1% 6 22.2% 1 20.0%0.0314

Hormone and binding protein study

Blood samples for hormone assays were drawn between 08:00 and 09:00 during the mid-follicular (7–10th day after menses onset) and mid-luteal (20–24th day) phases. E2, Pg, sex hormone binding globulin (SHBG), and albumin were assayed in duplicate by routine laboratory methods. Analytic sensitivities were 55 pmol/L for E2, 0.3 nmol/L for Pg, and 0.2 nmol/L for SHBG.

Free E2 (fE2) concentrations were calculated by the formula: inline imageinline image (Södergard et al., 1982). Results obtained by applying this formula correlate well with the direct measurements of fE2 fraction by equilibrium dialysis, and are more accurate than the calculation of the fE2 index by dividing E2 for SHBG concentrations. Molar E2/Pg and fE2/Pg ratios were also calculated.

Data analysis

Data were represented by the means and their lower and upper 95% confidence limits (95% CIs), or the median and ranges for seizure number in the previous 6-month period. All analyses were run by using version 12.0 of the SPSS statistical package, for Windows (Chicago, IL, U.S.A.).

One-way analyses of variance (ANOVAs) were used to test for statistical differences in normally distributed variables, and Kruskal-Wallis test for those with nonparametric distribution. The Bonferroni-Dunn’s post hoc test was applied when two or more mean data were compared as appropriate. Chi-square (χ2) and Fischer’s exact test were used for categorical data, where appropriate.

Bivariate correlations between ordered variables were analyzed using the Spearman correlation analysis. E2/Pg and fE2/Pg ratios were processed after log transformation of the values. A p-value < 0.05 (two-sided) was considered indicative of a statistically significant difference.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Clinical and demographic data

Women with more frequent seizures (SFS 3–4) presented with a significantly earlier disease onset, and longer disease duration. The prevalence of AED polytherapies progressively increased from 23.1% in the SFS 1 up to 80% in SFS 4 group (Table 2).

No significant SFS-related difference was found in the distribution of anovulatory cycles as assessed by luteal Pg levels < 6.36 nmol/L, but cases with luteal Pg levels between 6.37 and 15.8 nmol/L were significantly more frequent in the SFS 3–4 groups (Table 2).

Hormonal data

E2, Pg, and fE2 levels in both ovarian phases, and follicular fE2/Pg ratios were significantly lower in PE than in healthy controls. Patients had significantly higher SHBG and lower albumin concentrations (Table 3).

Table 3.   Mean and 95% CI (confidence interval) of sex hormone binding protein, free estradiol (fE2) concentrations and related ratios during the follicular and luteal phases in 72 women with partial epilepsy (PE), and in 30 healthy controls
 PE patients (n = 72)Healthy women (n = 30)p-value
  1. Statistical analysis by analysis of variance (ANOVA).

  2. Pg, progesterone; SHBG, sex hormone binding globulin.

E2 follicular pmol/L122.9 102.7–143.0188.8 171.2–206.3<0.0001
E2 luteal pmol/L309.3 102.7–349.6493.6 441.9–545.4<0.0001
Pg follicular nmol/L  1.1   0.9–1.3  1.5   1.3–1.70.0054
Pg luteal nmol/L 20.4  17.1–23.7 32.6  28.0–37.20.0002
SHBG follicular nmol/L 68.2  59.0–77.4 44.9  42.1–47.70.0026
SHBG luteal nmol/L 69.2  60.2–78.1 47.7  44.0–51.30.0096
fE2 follicular pmol/L  0.72   0.59–0.84  1.38   1.22–1.550.0001
fE2 luteal pmol/L  2.08   1.59–2.57  3.43   2.98–3.880.0057
E2/Pg follicular 19.1  12.1–26.1 14.6  12.3–16.9NS
E2/Pg luteal  4.9   1.5–8.4  1.7   1.4–1.9NS
fE2/Pg follicular  8.3   6.1–10.6 10.8   9.1–12.50.0010
fE2/Pg luteal 36.1   7.2–64.9 11.6   9.6–13.5NS
Albumin g/L 44.0  43.2–44.8 47.3  45.7–48.90.0004

In women with PE divided on the basis of their SFS versus healthy controls, statistical significance for hormonal and binding protein levels were confirmed, but no significant differences among SFS groups were present (Table 4).

Table 4.   Mean and 95% CI of sex hormone binding protein, fE2 (free estradiol) concentrations and related ratios during the follicular and luteal phases in 72 women with partial epilepsy (PE) divided according to their seizure frequency scores (SFSs)
Sex hormonesSFS 1 (n = 26)SFS 2 (n = 14)SFS 3 (n = 27)SFS 4 (n = 5)SFS 1–4 versus healthy controls p
  1. Statistical analysis by the Kruskal-Wallis test. Letters in apex show significant differences (p < 0.05) versus healthy controls at the Bonferroni-Dunn’s post hoc test.

  2. Pg, progesterone; SHBG, sex hormone binding globulin.

E2 follicular pmol/L116.0  91.6–140.4130.3  79.2–181.5135.4  92.7–178.1 69.9a  14.6–125.2<0.0001
E2 luteal pmol/L314.7a 256.3–373.2286.4a 177.7–395.0315.8a 237.6–394.0310.7a 145.1–476.3<0.0001
Pg follicular nmol/L  1.2   0.9–1.5  1.0   0.7–1.4  0.9a   0.7–1.2  1.7   0.6–2.70.0012
Pg luteal nmol/L 23.3  17.6–28.9 23.2  13.3–33.1 16.6a  11.6–21.6 18.8a   5.7–21.90.0013
SHBG follicular nmol/L 62.3a  46.4–78.2 59.0a  32.3–85.6 79.9a  65.7–94.1 61.4a  29.7–93.10.0040
SHBG luteal nmol/L 60.7a  46.5–74.9 59.9a  33.7–86.2 81.4a  67.2–95.6 73.0a  29.4–116.60.0079
fE2 follicular pmol/L  0.83a   0.57–1.10  0.84a   0.54–1.14  0.60a   0.45–0.74  0.42a   0.03–0.820.0001
fE2 luteal pmol/L  2.50   1.51–3.50  2.14   1.18–3.10  1.71a   0.91–2.52  1.75a   0.25–3.250.0001
E2/Pg follicular 22.4   3.6–41.1 15.9   7.8–24.0 20.2  14.3–26.1  5.2a   0.2–10.10.0358
E2/Pg luteal  4.6   0.4–8.9  2.5   0.7–4.4  7.1 −1.6–15.8  1.9   1.0–2.9NS
fE2/Pg follicular 11.0   5.6–16.5  8.8   4.1–13.5 5.4a  3.2–7.6  8.8   4.9–12.70.0023
fE2/Pg luteal 59.9 −17.8–137.7 18.7   4.9–32.627.2  1.4–52.9  8.9   6.6–11.1NS
Albumin g/L 44.7  43.6–45.8 44.9  42.7–47.143.3a 42.0–44.7 41.8a  36.6–47.10.0017

However, when hormonal data in SFS groups were further analyzed, it emerged that the difference versus healthy controls in follicular E2, follicular and luteal Pg, luteal fE2, follicular E2/Pg and fE2/Pg, and albumin levels was due to significant (p < 0.05) changes occurring in patients included in SFS 3 and/or SFS 4 groups, but not in those with absent or rarer seizures (SFS 1–2) (Table 4).

No significant differences were found between hormone data in patients on AED monotherapy versus those on AED polytherapy (p > 0.05). Luteal Pg levels were lower in the AED polytherapy (17.3 nmol/L; 95% CI 11.4–23.3) than in the monotherapy group (22.3 nmol/L; 95% CI 18.4–26.3), but the difference was at the limit of statistical significance (p = 0.0811).

Correlations between clinical, biometric, and hormonal data

SFS correlated inversely with age at disease onset (p = 0.0007), and directly with disease duration (p < 0.0001). Moreover, disease duration was correlated directly with age (p < 0.0001), and inversely with age at disease onset (p < 0.0001). Free E2 correlated inversely with SFS (p = 0.0005) during the follicular phase and with disease duration (p = 0.0066) in the luteal one. Finally, an inverse correlation was found between SHBG levels and body weight (p = 0.0211 for the follicular phase; p = 0.0286 for the luteal phase).


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

In this study, we have evaluated consecutive outpatients with AED-treated PE and attempted to correlate ovary function and sex steroid data with seizure frequency and other clinical parameters.

Sex steroid levels in both ovarian phases were significantly lowered in women with PE compared to sex- and age-matched healthy controls. Changes in reproductive hormone levels during the ovarian cycle have been reported extensively in large groups of patients with various epilepsy syndromic types (Murialdo et al., 1997, 1998; Stoffel-Wagner et al., 1998; Isojärvi et al., 2005; Löfgren et al., 2007;Galimberti et al., 2009), but less frequently in selected cohorts including only PE subjects (Stoffel-Wagner et al., 1998; Herzog & Friedman, 2001).

Interictal discharges and generalized or partial seizures may impact reproductive function at different levels by affecting gonadotropin pulsatile secretion, hypothalamic–pituitary control on the gonadal function, and sex hormone production (Drislane et al., 1994; Nappi et al., 1994; Cummings et al., 1995; Murialdo et al., 1997; Bauer et al., 1998; Bilo et al., 2001; Herzog, 2008a). In temporal lobe epilepsy (TLE), limbic discharge is held responsible for changes in the release of hypothalamic hormones, which regulate gonadotropin secretion and ovary function (Cummings et al., 1995; Herzog et al., 2003; Herzog, 2008a). However, it should be stressed that we have found no difference in sex hormone parameters between seizure-free cases (SFS 1) and those with occasional or more frequent crises. Therefore, the presence of epilepsy rather than seizure frequency seems to be relevant in inducing reproductive dysfunction and sex hormone changes in females with AED-treated PE.

Moreover, most of the AEDs considered in this study induce the activity of the cytochrome P450 (CYP) oxidative system and activate the synthesis of SHBG by the hepatocyte, thereby interfering with steroid metabolism and bioavailability (Perucca et al., 1984; Murialdo et al., 1998; Morrell et al., 2001; Isojärvi et al., 2005; Isojärvi, 2008). In addition, we have found an inverse correlation between body weight and SHBG levels. This aspect should be carefully pondered in patients with epilepsy because changes in SHBG levels and sex hormone free fractions must be considered as one of the markers of obesity and cardiovascular disease (Tchernof & Despres, 2000).

The most intriguing finding in this study concerned the relationship between epilepsy severity and sex hormone levels, which were different from those in healthy controls in patients with more frequent seizures but not when seizures were rarer or absent. An earlier onset and longer duration of the disease, in addition to a larger prevalence of AED polytherapies, characterized the groups with higher SFS. Younger age and longer AED treatment were also found to be predictors of reproductive endocrine disorders in women with epilepsy taking VPA (Isojärvi et al., 2005).

Neither total E2 nor Pg levels in our patients correlated significantly with clinical parameters, but an inverse relationship was found between fE2 levels and SFS during the follicular phase, and with disease duration in the luteal one. PE women with more frequent seizures showed higher mean SHBG levels than those in SFS 1–2 groups, but the differences were not statistically significant because of the large dispersion of values. The finding may be attributed to the longer duration of AED treatments and the more frequent use of enzyme-inducing AEDs in subjects with more severe seizures. Such changes in serum SHBG along with lowered total E2 levels implied a decrease in the biologically active fE2 titers in women with more frequent seizures and longer disease duration. Therefore, the impairment of E2 biologic activity due to the decrease of its free fraction, rather than changes in total E2, may be ascribed to the seizure frequency, disease severity, and the stimulating effect of AED on SHBG synthesis.

We did not analyze the effects of single AEDs on sex hormones in our patients because of the great variety of pharmacologic treatments and the small number of cases in some groups. Most of our PE women were treated with traditional enzyme-inducing AEDs, such as PB and CBZ, whereas VPA, which is more frequently associated with reproductive dysfunction (Isojärvi, 2008), was administered in only a negligible number of subjects.

Estrogens are produced from androgens through the activity of aromatases, and the inhibition of these enzymes by some AEDs may involve a decreased production of E2, mainly when polytherapies with multiple interfering drugs are administered. An inhibition of the aromatase complex (CYP19) by most of the AEDs used in our patients, but not by CBZ and PRM, has been demonstrated in vitro (Jacobsen et al., 2008), and possibly also occurs in vivo.

The majority of clinical studies used endometrial biopsy as the standard to diagnose luteal phase defects. However, in common practice, measurements of serum Pg may be considered as a reliable means of identifying anovulation or luteal dysfunction. Although the best correlation with endometrial histology is achieved by repeated Pg assays obtained during days 5–9 after ovulation, the current study was designed with only one sample in the mid-luteal phase. Pg levels obtained in this way are clearly an indirect parameter of the menstrual cycle investigation. With this limitation taken into account, anovulatory cycles were found in 23.6% and a luteal phase defect with reduced Pg levels (6.37–15.8 nmol/L) in a further 15.3% of our PE women, so that an overall cumulative frequency of 38.9% of patients had serum Pg levels lower than 15.8 nmol/L. This limit has been considered by other authors as indicative of anovulation when combined with basal temperature monitoring. Our prevalence findings with respect to anovulation and luteal phase defect in menstruating women with PE are similar to those reported in the literature (Cummings et al., 1995; Murialdo et al., 1997; Bauer et al., 1998; Herzog & Friedman, 2001; Morrell et al., 2002; Herzog et al., 2004; Herzog, 2006).

These prevalences appear to be higher than that currently quoted in the general population, ranging between 8.0 and 10.9% (Abdulla et al., 1983); but the actual frequency of the luteal phase failure in this form of epilepsy remains uncertain, being possibly influenced by case selection, investigational procedures, and methodologic approaches.

The seizure elicitation by the physiologic surge of sex steroids during the luteal phase and by the subsequent premenstrual Pg drop has been emphasized, so that catamenial seizures might be more frequent in women with Pg surge indicative of ovulatory cycles than in anovulatory ones (Bauer et al., 1998).

The impairment of Pg secretion during the luteal phase we observed may be only in part ascribed to AED regimens. Anovulation has been reported to be more frequent in women with TLE taking polytherapy than in those receiving a single AED, but also in this instance the difference was not statistically significant and appeared to be unaffected by seizure frequency (Cummings et al., 1995).

Lowered sex hormone levels in epilepsy may be, in part, accounted for by the AED therapy, which involves a decrease in estrogen production and conversion from androgens in liver and peripheral tissues. In addition, both seizures and AEDs may affect hormone secretions and gonadotropin ovulatory surge by directly targeting a number of anatomic substrates, including the limbic system, hypothalamus, pituitary, and gonads (Herzog, 2008a).

Neuroexcitatory effects of estrogens and anticonvulsant properties exerted by progestins have been quoted previously. However, in our PE patients with more frequent seizures, E2 and fE2 levels were rather decreased, and E2/Pg or fE2/Pg ratios did not explain higher seizure rates. Therefore, actual circulating E2 levels did not appear to be correlated with an increased seizure occurrence despite the well-known effects of estrogens on neuron excitability largely supported by in vitro evidence. But hormone changes potentially explaining an enhanced seizure frequency rather concerned luteal Pg levels, in view of the anti-epileptogenic activity attributed to progestins, in part due to their actions on γ-aminobutyric acid (GABA)A receptor–mediated chloride conductance (Bäckström, 1976; Beyenburg et al., 2001).

E2 and Pg exert favorable biologic and metabolic effects in females, and these aspects should be comprehensively assessed for an optimal evaluation of women with epilepsy. Menstrual disorders and sex hormone levels should be monitored closely in these patients to prevent adverse effects of reproductive dysfunction and ovarian failure, mainly in women with more frequent seizures often exposed to AED polytherapies.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

This study was supported by grants from the Ministry of Health to the IRCCS ‘C. Mondino Institute of Neurology’ Foundation (RC 2006). We are grateful to Maria Vittoria Gianelli, Ph.D., Faculty of Medicine – University of Genova, for the linguistic revision of the manuscript, and to Rossella E. Nappi, M.D., Ph.D., for her useful advice about gynecologic concerns.

We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Disclosure: None of the authors has any conflict of interest to disclose.


  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References