1
Prevalence of Endocrine and Metabolic Disorders in Subjects with Erectile Dysfunction: A Comparative Study Elisa Maseroli, MD,* Giovanni Corona, MD, PhD,† Giulia Rastrelli, MD, PhD,* Francesco Lotti, MD,* Sarah Cipriani, MD,* Gianni Forti, MD,‡ Edoardo Mannucci, MD,§ and Mario Maggi, MD* *Sexual Medicine and Andrology Unit, Department of Experimental, Clinical, and Biomedical Sciences, University of Florence, Florence, Italy; †Endocrinology Unit, Medical Department, Azienda Usl, Maggiore-Bellaria Hospital, Bologna, Italy; ‡Endocrinology Unit, Department of Experimental, Clinical, and Biomedical Sciences, University of Florence, Florence, Italy; §Diabetes Agency, Careggi Hospital, Florence, Italy DOI: 10.1111/jsm.12832
ABSTRACT
Introduction. Alterations of gonadal, thyroid, and pituitary hormones, along with metabolic disorders, might be involved in causing erectile dysfunction (ED). Aim. The prevalence of endocrine abnormalities in two different cohorts from the general and the symptomatic populations of Florence was compared. Methods. The first group is a general population sample derived from a Florentine spin-off of the European Male Aging Study (EMAS cohort; n = 202); the second group is a series of n = 3,847 patients attending our clinic for ED (UNIFI cohort). Results. Both primary and secondary hypogonadism were more often observed in the UNIFI than in the EMAS cohort (2.8 vs. 0%; P < 0.05 and 18.9 vs. 8%; P < 0.001, respectively). However, only the second association retained statistical significance after adjusting for age. Compensated hypogonadism was more common in the EMAS cohort (4.4 vs. 8.1%; P < 0.05). No statistically significant difference in the prevalence of overt thyroid disorders was observed. Conversely, subclinical hyperthyroidism was more prevalent in the EMAS cohort (2 vs. 4.1%, P < 0.05). No significant difference in the prevalence of hyperprolactinemia was detected, while the prevalence of hypoprolactinemia was significantly higher in the UNIFI than in the EMAS cohort (28.2% vs. 17.8%, P = 0.001), even after the adjustment for age, BMI, and testosterone (P = 0.001). Central obesity (waist ≥102 cm), impaired fasting glucose (IFG), and type 2 diabetes mellitus (T2DM) were more often detected in UNIFI patients (31.7 vs. 22.8%, P < 0.05; 44.5 vs. 33.3%, P < 0.05; 20.1% vs. 1.0%, P < 0.001 in the UNIFI and EMAS cohort, respectively), even after adjusting for age. In contrast, the prevalence of overweight and obesity did not differ between the two groups. Conclusion. T2DM, IFG, central obesity, secondary hypogonadism, and hypoprolactinemia are more frequent in subjects consulting for ED than in the general population of the same geographic area. Our data suggest that these conditions could play a central role in determining consultation for ED. Maseroli E, Corona G, Rastrelli G, Lotti F, Cipriani S, Forti G, Mannucci E, and Maggi M. Prevalence of endocrine and metabolic disorders in subjects with erectile dysfunction: A comparative study. J Sex Med **;**:**–**. Key Words. Hypogonadism; Hormones; Metabolic Syndrome
Introduction
S
exual activity recapitulates a complex array of behaviors finally bringing together two separate partners, in the majority of cases of an oppo© 2015 International Society for Sexual Medicine
site gender, communicating their reciprocal sexual interest for reproduction or for some expression of intimacy. Hence, communication is the cornerstone of sexual activity. There are several languages used for sexual communication, but the J Sex Med **;**:**–**
2 most common one is hormones, messages sent from dedicated glands into the bloodstream for recipient targets. Therefore, when the cascade of hormonal activity is impaired, such as in the case of endocrine disorders, investigating sexual dysfunction is mandatory. Hormone alterations can be involved in determining erectile dysfunction (ED), or they can be the consequence of ED, often with a bidirectional interconnection [1–7]. Many endocrine disorders, or their treatments, are characterized by several alterations in sexual response. Accordingly, ED is considered the highest incident endocrine disorder in males in the general U.S. population [8]. Despite this evidence, there are no accurate estimates of the prevalence of endocrine-related disorders in sexual dysfunction in random, population-based, samples; in fact, most reported series come from selected clinic populations [9]. In addition, some of the available studies focus on ED, whereas others involve a wider range of sexual dysfunctions, some of which could have a relationship with endocrine diseases not linked to ED. El-Sakka et al., in a sample of 1,248 patients with sexual dysfunction, reported an overall prevalence of endocrine disorders of about 25% [10]. In particular, the most frequent endocrinopathies were hypogonadism (15%), hyperprolactinemia (13.7%), and hypothyroidism (3.1%). Unsuspected hypothyroidism was also detected in up to 6% of 600 men with ED [11], whereas hyperthyroidism was less often associated with ED [6,12]. The prevalence of hyperprolactinemia in male subjects with sexual dysfunction is quite variable, ranging from more than 13% (milder forms) to less than 1% (more severe forms) [13,14]. The prevalence of hypogonadism in ED subjects ranges from less than 10% up to more than 30%, varying according to the cutoff value adopted for the diagnosis [1,2,15,16]. Despite this evidence, the widespread screening for hypogonadism in ED subjects and the role of testosterone (T) supplementation as a possible treatment still remain questionable [15,16]. Obesity and type 2 diabetes mellitus (T2DM) are common metabolic disorders. Subjects with ED represent a population overloaded with these conditions, which contribute to increasing the overall high cardiovascular (CV) risk associated with ED [17–19]. Endocrine disorders in ED are usually considered easy to diagnose and often remediable [20]; however, as just mentioned, they frequently receive little attention. It should be recognized that the J Sex Med **;**:**–**
Maseroli et al. assessment of endocrine disorders can uncover serious diseases, such as pituitary tumors, or can help in stratifying the CV risk of ED patients [14,15]. The aim of the present study is to overcome the lack of information in current literature by comparing the prevalence of several endocrinopathies in the same geographic area. We used data from a general population sample (as derived from a spinoff of the European Male Aging Study [EMAS], Florence center database: EMAS cohort) and those from a large consecutive series of patients seeking medical care for ED (University of Florence Study, UNIFI cohort). Hormones in the two cohorts were studied using the same analytical methods.
Materials and Methods
This is a cross-sectional study involving two different cohorts of subjects.
EMAS Cohort (General Population) EMAS is a prospective population-based cohort study of male ageing conducted on noninstitutionalized men aged 40–79 years, recruited from eight European centers in the period 2003– 2004. Details regarding recruitment, response rates, and assessments have been described in detail elsewhere [21–23]. Briefly, stratified random sampling was used with the aim of recruiting equal numbers of men into four 10-year age bands (40– 49, 50–59, 60–69, and 70–79 years). There were no specific exclusion criteria apart from subjects having to provide written, informed consent. Subjects were invited by letter to attend a screening at a local clinic. In Florence, at the time of study start-up, there were 112,689 men aged 40–79 [24]; a random sample of 700 individuals were invited, and 433 subjects responded and were enrolled, accounting for a participation rate of 61.5% [25]. As a high prevalence of sexual dysfunction was detected in the EMAS population, we excluded those subjects who reported at least one of the following symptoms: ED, low libido, or decreased morning erections, as assessed by the EMAS Sexual Function Questionnaire [23], categorized according to previously reported criteria [26,27]. After excluding subjects having at least one sexual dysfunction symptom, 46.5% (n = 202) of subjects were included in the present analysis. The sociodemographical and clinical characteristics of the sample are reported in Table 1.
3
Endocrine and Metabolic Disorders in Subjects with Erectile Dysfunction Table 1
Characteristics of the sample
Characteristics Age (years) Clinical and laboratory parameters BMI (kg/m2) Waist circumference (cm) Glycemia (mmol/L) LH (mU/L) Total testosterone (nmol/L) PRL (mU/L) TSH (mU/L) FT4 (pmol/L)
EMAS cohort (n = 202)
UNIFI cohort (n = 3,847)
55.25 ± 9.8
60.13 ± 10.9
26.9 ± 3.6 95.8 ± 9.4 5.3 (4.9–5.8) 4.72 (3.5–6.1) 16.3 ± 4.49 164.95 (124.3–222) 1.39 (0.9–1.9) 15.94 (14.5–17.4)
27.1 ± 4.2 99.0 ± 10.8 5.4 (4.9–6.2) 3.8 (2.6–5.6) 15.2 ± 6.2 148.0 (107.0–216.0) 1.4 (0.9–1.9) 13.9 (11.9–15.9)
Data are expressed as mean ± standard deviation when normally distributed, as median (quartiles) when not normally distributed, and as percentages when categorical BMI = body mass index. LH = luteinizing hormone. FSH = follicle stimulating hormone. PRL = prolactin. TSH = thyroid-stimulating hormone. FT4 = free thyroxin
UNIFI Cohort (Patients with Sexual Dysfunction) An unselected consecutive series of male patients attending our Andrology and Sexual Medicine Outpatient Clinic for the first time for ED were retrospectively studied at the University of Florence (UNIFI cohort). Subjects aged 40–79 years (n = 3,847) were selected. The sociodemographical and clinical characteristics of the sample are reported in Table 1. All patients enrolled underwent the usual diagnostic protocol applied to newly referred subjects at the Andrology Outpatient Clinic as previously reported [28–32]. All the data provided were collected as part of the routine clinical procedure. Clinical and Biochemical Analyses All patients underwent a complete physical examination in both studies, including blood pressure, weight, height, and waist circumference measurement. Blood samples were drawn in the morning, after an overnight fast, for determination of blood glucose (by glucose oxidase method), total T, luteinizing hormone (LH), prolactin (PRL), and thyroid stimulating hormone (TSH) (by electrochemiluminescent method). In the UNIFI sample, free thyroxin (FT4) was measured only when TSH values were outside of the reference ranges (0.4–4.5 mU/L), whereas in the EMAS cohort, it was systematically detected as previously reported [6]. In the EMAS cohort, T was retested by gas chromatography–mass spectrometry [33]. Statistical Analysis Differences in the prevalence of endocrine/ metabolic diseases between EMAS and UNIFI cohort were assessed by χ2-test. Binary logistic regression was applied for multivariate analysis
using Hosmer–Lemeshow test for assessing the goodness of fit of the model. Each analysis was firstly adjusted for age; further adjustments were made with proper confounding factors potentially interfering with the prevalence of each endocrine or metabolic disorder. In particular, when the difference in prevalence of T2DM between the two groups was assessed, waist circumference was entered into the analysis as a further covariate; similarly, both total T and body mass index (BMI) were entered into the model when evaluating the difference in prevalence of hypoprolactinemia and central obesity. As mentioned before, our sample size is made up of 202 subjects from the EMAS cohort and 3,847 subjects from the UNIFI cohort. Conversely from what was expected on the basis of epidemiological studies [19], we failed to find a significant difference in prevalence of obesity between the two study groups. Considering that the prevalence of obesity was 16.2% in the EMAS cohort, we would be able to detect a significant difference between our two groups if the prevalence of obesity among the UNIFI cohort had been at least 24%. We fixed a power of 80% and a type I error probability associated with the null hypothesis (prevalence of obesity in EMAS and UNIFI cohort are equal) of 0.05. The null hypothesis was evaluated by chi-squared statistic. All statistical analyses were performed using spss (Statistical Package for the Social Sciences, Chicago, IL, USA) for Windows 20.0.0.
Results
Figure 1 shows the prevalence of the investigated endocrine (panel A) and metabolic (panel B) disorders in the EMAS and UNIFI cohorts. J Sex Med **;**:**–**
4
Maseroli et al.
Figure 1 Prevalence (%) of endocrine (A) and metabolic (B) abnormalities in the Florentine subjects of the European Male Aging Study (EMAS cohort) (n = 202) and in a consecutive series of n = 3,847 male patients attending our clinic for erectile dysfunction at the University of Florence (UNIFI cohort). NS = not significant. *P < 0.05, **P < 0.001
Hypogonadism Both primary (total T < 10.5 nmol/L and LH >9.4 U/L) and secondary hypogonadism (total T < 10.5 nmol/L and LH ≤9.4 U/L) were more often observed in the UNIFI than in the EMAS cohort (2.8 vs. 0%; P < 0.05 and 18.9 vs. 8%; P < 0.001, respectively). However, only the second association retained statistical significance after adjusting for age (Figure 2). Compensated hypogonadism (total T ≥ 10.5 nmol/L and LH > 9.4 U/L) was more common in the EMAS cohort (4.4 vs. 8.1%; P < 0.05), even after adjusting for age (Figure 2).
cohorts, including overt hypothyroidism (TSH values > 5.5 mU/L and FT4 < 11.5 pmol/L) or hyperthyroidism (TSH levels < 0.35 mU/L and FT4 values > 23 pmol/L), was observed. Conversely, subclinical hyperthyroidism (TSH levels < 0.35 mU/L and FT4 within the normal range), but not subclinical hypothyroidism (TSH levels between 5.5 and 10 mU/L and FT4 within the normal range), was more prevalent in the EMAS cohort (2 vs. 4.1%, P < 0.05; Figure 1), even after adjusting for age (Figure 2).
Hypo and Hyperthyroidism No statistically significant difference in the prevalence of overt thyroid disorders in the two
Hypo and Hyperprolactinemia No significant difference in the prevalence of hyperprolactinemia, either mild (PRL >420 mU/L
J Sex Med **;**:**–**
Endocrine and Metabolic Disorders in Subjects with Erectile Dysfunction
5
Figure 2 Age-adjusted odd ratio for endocrine and metabolic abnormalities in the Florentine cohort (n = 202) of the European Male Aging Study (EMAS cohort) in comparison with a consecutive series of n = 3,847 male patients attending our clinic for erectile dysfunction at the University of Florence (UNIFI cohort)
or >20 ng/mL) or severe (PRL > 735 mU/L or >35 ng/mL), was detected between the two groups (Figure 1). When hypoprolactinemia (PRL < 113 mU/L; 5 ng/mL) was investigated, its prevalence was significantly higher in the UNIFI than in the EMAS cohort (28.2% vs. 17.8%, P = 0.001; Figure 1). The latter finding was confirmed even after the adjustment for age (Figure 2). In addition, the introduction in a logistic model of BMI and T levels did not modify the latter association (odds ratio [OR] = 1.95 [1.32–2.87]; P = 0.001].
Metabolic Disorders When metabolic disorders were analyzed, central obesity (waist circumference ≥ 102 cm), impaired fasting glucose (IFG; fasting glucose levels >100 mg/dl), and T2DM were more often detected in UNIFI patients (31.7 vs. 22.8%, P < 0.05; 44.5 vs. 33.3%, P < 0.05; 20.1% vs. 1.0%, P < 0.001 in the UNIFI and EMAS cohort, respectively). These differences were confirmed in an age-adjusted model (Figure 2). Interestingly, the increased prevalence of central obesity in the UNIFI cohort was confirmed even when T levels and BMI were entered into the model as covariates (OR = 2.24 [1.31–3.82]; P < 0.05). In addition, the increased prevalence of T2DM in UNIFI cohort retained statistical significance even after adjusting for age and waist (OR 22.94 [5.67–92.85]; P < 0.001). In contrast, the prevalence of overweight (BMI ≥ 25 and
Prevalence of Endocrine and Metabolic Disorders in Subjects with Erectile Dysfunction: A Comparative Study Elisa Maseroli, MD,* Giovanni Corona, MD, PhD,† Giulia Rastrelli, MD, PhD,* Francesco Lotti, MD,* Sarah Cipriani, MD,* Gianni Forti, MD,‡ Edoardo Mannucci, MD,§ and Mario Maggi, MD* *Sexual Medicine and Andrology Unit, Department of Experimental, Clinical, and Biomedical Sciences, University of Florence, Florence, Italy; †Endocrinology Unit, Medical Department, Azienda Usl, Maggiore-Bellaria Hospital, Bologna, Italy; ‡Endocrinology Unit, Department of Experimental, Clinical, and Biomedical Sciences, University of Florence, Florence, Italy; §Diabetes Agency, Careggi Hospital, Florence, Italy DOI: 10.1111/jsm.12832
ABSTRACT
Introduction. Alterations of gonadal, thyroid, and pituitary hormones, along with metabolic disorders, might be involved in causing erectile dysfunction (ED). Aim. The prevalence of endocrine abnormalities in two different cohorts from the general and the symptomatic populations of Florence was compared. Methods. The first group is a general population sample derived from a Florentine spin-off of the European Male Aging Study (EMAS cohort; n = 202); the second group is a series of n = 3,847 patients attending our clinic for ED (UNIFI cohort). Results. Both primary and secondary hypogonadism were more often observed in the UNIFI than in the EMAS cohort (2.8 vs. 0%; P < 0.05 and 18.9 vs. 8%; P < 0.001, respectively). However, only the second association retained statistical significance after adjusting for age. Compensated hypogonadism was more common in the EMAS cohort (4.4 vs. 8.1%; P < 0.05). No statistically significant difference in the prevalence of overt thyroid disorders was observed. Conversely, subclinical hyperthyroidism was more prevalent in the EMAS cohort (2 vs. 4.1%, P < 0.05). No significant difference in the prevalence of hyperprolactinemia was detected, while the prevalence of hypoprolactinemia was significantly higher in the UNIFI than in the EMAS cohort (28.2% vs. 17.8%, P = 0.001), even after the adjustment for age, BMI, and testosterone (P = 0.001). Central obesity (waist ≥102 cm), impaired fasting glucose (IFG), and type 2 diabetes mellitus (T2DM) were more often detected in UNIFI patients (31.7 vs. 22.8%, P < 0.05; 44.5 vs. 33.3%, P < 0.05; 20.1% vs. 1.0%, P < 0.001 in the UNIFI and EMAS cohort, respectively), even after adjusting for age. In contrast, the prevalence of overweight and obesity did not differ between the two groups. Conclusion. T2DM, IFG, central obesity, secondary hypogonadism, and hypoprolactinemia are more frequent in subjects consulting for ED than in the general population of the same geographic area. Our data suggest that these conditions could play a central role in determining consultation for ED. Maseroli E, Corona G, Rastrelli G, Lotti F, Cipriani S, Forti G, Mannucci E, and Maggi M. Prevalence of endocrine and metabolic disorders in subjects with erectile dysfunction: A comparative study. J Sex Med **;**:**–**. Key Words. Hypogonadism; Hormones; Metabolic Syndrome
Introduction
S
exual activity recapitulates a complex array of behaviors finally bringing together two separate partners, in the majority of cases of an oppo© 2015 International Society for Sexual Medicine
site gender, communicating their reciprocal sexual interest for reproduction or for some expression of intimacy. Hence, communication is the cornerstone of sexual activity. There are several languages used for sexual communication, but the J Sex Med **;**:**–**
2 most common one is hormones, messages sent from dedicated glands into the bloodstream for recipient targets. Therefore, when the cascade of hormonal activity is impaired, such as in the case of endocrine disorders, investigating sexual dysfunction is mandatory. Hormone alterations can be involved in determining erectile dysfunction (ED), or they can be the consequence of ED, often with a bidirectional interconnection [1–7]. Many endocrine disorders, or their treatments, are characterized by several alterations in sexual response. Accordingly, ED is considered the highest incident endocrine disorder in males in the general U.S. population [8]. Despite this evidence, there are no accurate estimates of the prevalence of endocrine-related disorders in sexual dysfunction in random, population-based, samples; in fact, most reported series come from selected clinic populations [9]. In addition, some of the available studies focus on ED, whereas others involve a wider range of sexual dysfunctions, some of which could have a relationship with endocrine diseases not linked to ED. El-Sakka et al., in a sample of 1,248 patients with sexual dysfunction, reported an overall prevalence of endocrine disorders of about 25% [10]. In particular, the most frequent endocrinopathies were hypogonadism (15%), hyperprolactinemia (13.7%), and hypothyroidism (3.1%). Unsuspected hypothyroidism was also detected in up to 6% of 600 men with ED [11], whereas hyperthyroidism was less often associated with ED [6,12]. The prevalence of hyperprolactinemia in male subjects with sexual dysfunction is quite variable, ranging from more than 13% (milder forms) to less than 1% (more severe forms) [13,14]. The prevalence of hypogonadism in ED subjects ranges from less than 10% up to more than 30%, varying according to the cutoff value adopted for the diagnosis [1,2,15,16]. Despite this evidence, the widespread screening for hypogonadism in ED subjects and the role of testosterone (T) supplementation as a possible treatment still remain questionable [15,16]. Obesity and type 2 diabetes mellitus (T2DM) are common metabolic disorders. Subjects with ED represent a population overloaded with these conditions, which contribute to increasing the overall high cardiovascular (CV) risk associated with ED [17–19]. Endocrine disorders in ED are usually considered easy to diagnose and often remediable [20]; however, as just mentioned, they frequently receive little attention. It should be recognized that the J Sex Med **;**:**–**
Maseroli et al. assessment of endocrine disorders can uncover serious diseases, such as pituitary tumors, or can help in stratifying the CV risk of ED patients [14,15]. The aim of the present study is to overcome the lack of information in current literature by comparing the prevalence of several endocrinopathies in the same geographic area. We used data from a general population sample (as derived from a spinoff of the European Male Aging Study [EMAS], Florence center database: EMAS cohort) and those from a large consecutive series of patients seeking medical care for ED (University of Florence Study, UNIFI cohort). Hormones in the two cohorts were studied using the same analytical methods.
Materials and Methods
This is a cross-sectional study involving two different cohorts of subjects.
EMAS Cohort (General Population) EMAS is a prospective population-based cohort study of male ageing conducted on noninstitutionalized men aged 40–79 years, recruited from eight European centers in the period 2003– 2004. Details regarding recruitment, response rates, and assessments have been described in detail elsewhere [21–23]. Briefly, stratified random sampling was used with the aim of recruiting equal numbers of men into four 10-year age bands (40– 49, 50–59, 60–69, and 70–79 years). There were no specific exclusion criteria apart from subjects having to provide written, informed consent. Subjects were invited by letter to attend a screening at a local clinic. In Florence, at the time of study start-up, there were 112,689 men aged 40–79 [24]; a random sample of 700 individuals were invited, and 433 subjects responded and were enrolled, accounting for a participation rate of 61.5% [25]. As a high prevalence of sexual dysfunction was detected in the EMAS population, we excluded those subjects who reported at least one of the following symptoms: ED, low libido, or decreased morning erections, as assessed by the EMAS Sexual Function Questionnaire [23], categorized according to previously reported criteria [26,27]. After excluding subjects having at least one sexual dysfunction symptom, 46.5% (n = 202) of subjects were included in the present analysis. The sociodemographical and clinical characteristics of the sample are reported in Table 1.
3
Endocrine and Metabolic Disorders in Subjects with Erectile Dysfunction Table 1
Characteristics of the sample
Characteristics Age (years) Clinical and laboratory parameters BMI (kg/m2) Waist circumference (cm) Glycemia (mmol/L) LH (mU/L) Total testosterone (nmol/L) PRL (mU/L) TSH (mU/L) FT4 (pmol/L)
EMAS cohort (n = 202)
UNIFI cohort (n = 3,847)
55.25 ± 9.8
60.13 ± 10.9
26.9 ± 3.6 95.8 ± 9.4 5.3 (4.9–5.8) 4.72 (3.5–6.1) 16.3 ± 4.49 164.95 (124.3–222) 1.39 (0.9–1.9) 15.94 (14.5–17.4)
27.1 ± 4.2 99.0 ± 10.8 5.4 (4.9–6.2) 3.8 (2.6–5.6) 15.2 ± 6.2 148.0 (107.0–216.0) 1.4 (0.9–1.9) 13.9 (11.9–15.9)
Data are expressed as mean ± standard deviation when normally distributed, as median (quartiles) when not normally distributed, and as percentages when categorical BMI = body mass index. LH = luteinizing hormone. FSH = follicle stimulating hormone. PRL = prolactin. TSH = thyroid-stimulating hormone. FT4 = free thyroxin
UNIFI Cohort (Patients with Sexual Dysfunction) An unselected consecutive series of male patients attending our Andrology and Sexual Medicine Outpatient Clinic for the first time for ED were retrospectively studied at the University of Florence (UNIFI cohort). Subjects aged 40–79 years (n = 3,847) were selected. The sociodemographical and clinical characteristics of the sample are reported in Table 1. All patients enrolled underwent the usual diagnostic protocol applied to newly referred subjects at the Andrology Outpatient Clinic as previously reported [28–32]. All the data provided were collected as part of the routine clinical procedure. Clinical and Biochemical Analyses All patients underwent a complete physical examination in both studies, including blood pressure, weight, height, and waist circumference measurement. Blood samples were drawn in the morning, after an overnight fast, for determination of blood glucose (by glucose oxidase method), total T, luteinizing hormone (LH), prolactin (PRL), and thyroid stimulating hormone (TSH) (by electrochemiluminescent method). In the UNIFI sample, free thyroxin (FT4) was measured only when TSH values were outside of the reference ranges (0.4–4.5 mU/L), whereas in the EMAS cohort, it was systematically detected as previously reported [6]. In the EMAS cohort, T was retested by gas chromatography–mass spectrometry [33]. Statistical Analysis Differences in the prevalence of endocrine/ metabolic diseases between EMAS and UNIFI cohort were assessed by χ2-test. Binary logistic regression was applied for multivariate analysis
using Hosmer–Lemeshow test for assessing the goodness of fit of the model. Each analysis was firstly adjusted for age; further adjustments were made with proper confounding factors potentially interfering with the prevalence of each endocrine or metabolic disorder. In particular, when the difference in prevalence of T2DM between the two groups was assessed, waist circumference was entered into the analysis as a further covariate; similarly, both total T and body mass index (BMI) were entered into the model when evaluating the difference in prevalence of hypoprolactinemia and central obesity. As mentioned before, our sample size is made up of 202 subjects from the EMAS cohort and 3,847 subjects from the UNIFI cohort. Conversely from what was expected on the basis of epidemiological studies [19], we failed to find a significant difference in prevalence of obesity between the two study groups. Considering that the prevalence of obesity was 16.2% in the EMAS cohort, we would be able to detect a significant difference between our two groups if the prevalence of obesity among the UNIFI cohort had been at least 24%. We fixed a power of 80% and a type I error probability associated with the null hypothesis (prevalence of obesity in EMAS and UNIFI cohort are equal) of 0.05. The null hypothesis was evaluated by chi-squared statistic. All statistical analyses were performed using spss (Statistical Package for the Social Sciences, Chicago, IL, USA) for Windows 20.0.0.
Results
Figure 1 shows the prevalence of the investigated endocrine (panel A) and metabolic (panel B) disorders in the EMAS and UNIFI cohorts. J Sex Med **;**:**–**
4
Maseroli et al.
Figure 1 Prevalence (%) of endocrine (A) and metabolic (B) abnormalities in the Florentine subjects of the European Male Aging Study (EMAS cohort) (n = 202) and in a consecutive series of n = 3,847 male patients attending our clinic for erectile dysfunction at the University of Florence (UNIFI cohort). NS = not significant. *P < 0.05, **P < 0.001
Hypogonadism Both primary (total T < 10.5 nmol/L and LH >9.4 U/L) and secondary hypogonadism (total T < 10.5 nmol/L and LH ≤9.4 U/L) were more often observed in the UNIFI than in the EMAS cohort (2.8 vs. 0%; P < 0.05 and 18.9 vs. 8%; P < 0.001, respectively). However, only the second association retained statistical significance after adjusting for age (Figure 2). Compensated hypogonadism (total T ≥ 10.5 nmol/L and LH > 9.4 U/L) was more common in the EMAS cohort (4.4 vs. 8.1%; P < 0.05), even after adjusting for age (Figure 2).
cohorts, including overt hypothyroidism (TSH values > 5.5 mU/L and FT4 < 11.5 pmol/L) or hyperthyroidism (TSH levels < 0.35 mU/L and FT4 values > 23 pmol/L), was observed. Conversely, subclinical hyperthyroidism (TSH levels < 0.35 mU/L and FT4 within the normal range), but not subclinical hypothyroidism (TSH levels between 5.5 and 10 mU/L and FT4 within the normal range), was more prevalent in the EMAS cohort (2 vs. 4.1%, P < 0.05; Figure 1), even after adjusting for age (Figure 2).
Hypo and Hyperthyroidism No statistically significant difference in the prevalence of overt thyroid disorders in the two
Hypo and Hyperprolactinemia No significant difference in the prevalence of hyperprolactinemia, either mild (PRL >420 mU/L
J Sex Med **;**:**–**
Endocrine and Metabolic Disorders in Subjects with Erectile Dysfunction
5
Figure 2 Age-adjusted odd ratio for endocrine and metabolic abnormalities in the Florentine cohort (n = 202) of the European Male Aging Study (EMAS cohort) in comparison with a consecutive series of n = 3,847 male patients attending our clinic for erectile dysfunction at the University of Florence (UNIFI cohort)
or >20 ng/mL) or severe (PRL > 735 mU/L or >35 ng/mL), was detected between the two groups (Figure 1). When hypoprolactinemia (PRL < 113 mU/L; 5 ng/mL) was investigated, its prevalence was significantly higher in the UNIFI than in the EMAS cohort (28.2% vs. 17.8%, P = 0.001; Figure 1). The latter finding was confirmed even after the adjustment for age (Figure 2). In addition, the introduction in a logistic model of BMI and T levels did not modify the latter association (odds ratio [OR] = 1.95 [1.32–2.87]; P = 0.001].
Metabolic Disorders When metabolic disorders were analyzed, central obesity (waist circumference ≥ 102 cm), impaired fasting glucose (IFG; fasting glucose levels >100 mg/dl), and T2DM were more often detected in UNIFI patients (31.7 vs. 22.8%, P < 0.05; 44.5 vs. 33.3%, P < 0.05; 20.1% vs. 1.0%, P < 0.001 in the UNIFI and EMAS cohort, respectively). These differences were confirmed in an age-adjusted model (Figure 2). Interestingly, the increased prevalence of central obesity in the UNIFI cohort was confirmed even when T levels and BMI were entered into the model as covariates (OR = 2.24 [1.31–3.82]; P < 0.05). In addition, the increased prevalence of T2DM in UNIFI cohort retained statistical significance even after adjusting for age and waist (OR 22.94 [5.67–92.85]; P < 0.001). In contrast, the prevalence of overweight (BMI ≥ 25 and
