The impact of age, BMI and sex hormone on aging males’ symptoms and the international index of erectile function scores

Abstract

Objective: To analyze the impact of age, BMI and sex hormone on aging males’ symptoms (AMS) and the 5-item version of the international index of erectile function (IIEF-5) scores in middle-aged and elderly Chinese men.

Methods: A population-based cross-sectional study was conducted in Jiashan County. A total of 969 men, aged between 40 and 80 years old, were admitted. Physical examination and the sex hormones were measured, and AMS and IIEF-5 scores were assessed.

Results: The oneway ANOVA analysis indicated older age groups had higher AMS total-scores, somatic and sexual sub-scores, and lower IIEF5 scores (all p < .01). Pairwise correlation (r_pairwise) analyses showed the significant associations between AMS and age or sex hormone (cFT, Bio-T, SHBG, and LH) levels, and similar for IIEF5. However, when age was adjusted, the correlation coefficients (r_partial) weakened, and correlation significance disappeared, except LH (for AMS: r_partial = 0.096, p = .009; for IIEF-5: r_partial =  −0.140, p = .001). Multiple linear regressions confirmed the influence of increased age and LH on the AMS and IIEF5 scores.

Conclusion: CFT, Bio-T and SHBG failed to yield any additional predicting information when age was adjusted. To improve the male reproductive health, future research should pay more attention on aging-related comorbidities and how to improve general wellness.

Keywords:

• Aging
• androgenic sex hormone
• erectile dysfunction
• aging males’ symptoms
• body mass index

Introduction

With an overall increase in life expectancy, aging is causing more severe public health issues globally. Reproductive health, as an important aspect of wellness, impacts on self-esteem, quality of life and interpersonal relationships [1]. How to delay aging on reproductive function and promote healthy aging, is still a thorny problem.

In China, aging males’ symptoms (AMS) scale and the international index of erectile function (IIEF-5) were widely used in clinical or scientific research for assessing Symptomatic Late-Onset Hypogonadism (SLOH) and Erectile Dysfunction (ED), respectively [2,3]. Erectile dysfunction (ED) is highly prevalent in hypogonadal men [4]. Although testosterone and other sex hormones were reported to be related to ED or AMS, androgen deficiency is responsible for ED or AMS only in a small subset of aging men [5]. Furthermore, various researches tested different hormones, testosterone, bioavailable testosterone (Bio-T), and or calculated free testosterone (cFT), and got inconsistent results [6–8]. It is imperative to make a comprehensive assessment on the relationship of sex hormone and reproductive health.

Materials and methods

Study population

In 2012, a cross-sectional study was performed in Jiashan County, Zhejiang province, which is a representative county of East China. Local male residents, aged 40 to 80 years old, were invited to volunteer to the study. Middle-aged and elderly men, who had history of psychotic, cognitive disorders, or took hormones, antiandrogen, or psychotropic agents, or had undergone pelvic surgical therapy, were excluded.

Clinical assessments

All the participants were asked to complete a uniform questionnaire, including demographic information, an abridged 5-item version of International Index of Erectile Function (IIEF-5) and Aging Males’ Symptoms (AMS) scale.

The AMS scale consists of 17 items, 5 for psychological, 7 for somatic and 5 for sexual subscores. Each item is scored as 1 to 5 points, and total score ranges from 17 to 85. The IIEF-5 includes 5 questions and each is scored on a rating scale of 0–5, for an overall score of 0–25 [9,10].

Endocrinology

All blood samples were drawn between 8:00 a.m. and 10:00 a.m. from overnight-fasting males to minimize the effects of diurnal variation. All hormonal assessments were conducted by the same laboratory, including total testosterone (TT), sex hormone-binding globulin (SHBG) and luteinizing hormone (LH). Serum level of calculated free testosterone (cFT) and bioavailable testosterone (Bio-T) was calculated according the equation [11].

Statistical analysis

The respondents were classified into four age groups: 40 to 50, 51 to 60, 61 to 70, and 71 years or older. Body mass index (BMI) was calculated and categorized into three groups: less than 18.5, 18.5 to 24, and 24 or more. Analysis of variance (ANOVA) was applied to compare the differences of means of AMS score or IIEF5 score among age groups or BMI groups.

Pearson’s correlation analysis was used to describe associations between the age, BMI, sex hormones and AMS or IIEF5. To eliminate the influence of multi-variable, partial correlation analysis was performed.

Multiple linear regression was applied to analyze the linear relationship among variables. Considering cFT and Bio-T were calculated from TT, SHBG and serum albumin, variance inflation factors (VIFs) were estimated in the multiple linear regressions. Strong associations among independent variables may induce large VIF value. VIF >10 indicates multi-collinearity, and then the linear regression model needs to be adjusted.

All statistics were calculated by using SAS9.1.3 package (SAS Institute, Cary, NC) and Stata 14 (Stata Corporation, College Station, TX).

Results

A total of 969 men, aged 56.4 ± 9.1 years old, were admitted to the data analysis. Means of BMI, TT, cFT, Bio-T, SHBG, LH, AMS score (psychological, somatic, sexual) and IIEF score were 23.7 kg/m², 4.1 ng/ml, 67.9 ng/l, 1.8 ng/ml, 43.3 nmol/l, 5.9 IU/l, 22.7 (5.2, 9.3, 8.2) and 17.2, respectively. The mean of each parameter was adjacent to the corresponding median (Table 1).

Table 1. Description of sex hormone levels, AMS and IIEF scores.

Variable	Min	Max	Median	Mean	Std.
Age	40	81	57	56.4	9.1
BMI (kg/m^2)	16. 2	34.4	23.7	23.7	2.8
TT (ng/ml)	0.1	11.1	3.9	4.1	1.1
cFT (ng/l)	11.4	142	66.6	67.9	16.5
Bio-T (ng/ml)	0.1	3.9	1.7	1.8	0.4
SHBG (nmol/l)	10.8	125.3	40.1	43.3	19.2
LH (IU/l)	0.1	33.3	5.0	5.9	3.6
AMS	17	62	21	22.7	6.7
Psychological	5	17	5	5.2	0.9
Somatic	7	24	8	9.3	3.1
Sexual	5	25	7	8.2	3.8
IIEF5	0	25	18	17.2	6.3

The oneway ANOVA analysis indicated total AMS scores and IIEF5 scores were significantly different among different age groups (all p < .01). Older age groups showed higher AMS scores and lower IIEF5 scores. When the AMS sub-scores were assessed, significant differences were found in the somatic and sexual sub-scores (all p < .01). However, there was no significant difference among the groups in terms of psychological sub-scores (p = .38) (Figure 1). No significant difference was found in the AMS scores or IIEF5 scores among different BMI groups (Figure 2).

Figure 1. Averages of AMS and IIEF5 scores stratified by age group.

Figure 2. Averages of AMS and IIEF5 scores stratified by BMI group.

Pairwise correlation (pairwise r) analyses showed the significant associations between AMS and age or sex hormone (cFT, Bio-T, SHBG, and LH) levels, and similar associations between IIEF5 and age or sex hormone. No correlation was found between AMS or IIEF5 and TT. However, when the age variable was adjusted, the correlation coefficients (partial r) between AMS and sex hormone (cFT, Bio-T and SHBG) or between IIEF5 and sex hormone weakened, and correlation significance disappeared, except LH. In the partial correlation analyses, age (r_partial = 0.353, p < .001) and LH (r_partial = 0.096, p = .009) showed positive and significant relationship with AMS, and negative significant relationship with IIEF (age: r_partial = −0.470, p < .001; LH: r_partial = −0.140, p = .001) (Table 2).

Table 2. Pairwise correlation and partial correlation between variables.

	rpairwise	p	rpartial	p
AMS-
Age	0.436	<.001	0.353	<.001
BMI	−0.045	.167	−0.012	.741
TT	0.047	.146	−0.029	.437
cFT	−0.153	<.001	0.049	.180
Bio-T	−0.189	<.001	−0.015	.683
SHBG	0.240	<.001	0.027	.470
LH	0.242	<.001	0.096	.009
IIEF5-
Age	−0.554	<.001	−0.470	<.001
BMI	0.063	.052	0.017	.655
TT	−0.065	.044	0.036	.329
cFT	0.178	<.001	−0.017	.640
Bio-T	0.210	<.001	−0.010	.783
SHBG	−0.288	<.001	−0.024	.523
LH	−0.308	<.001	−0.140	.001

Multiple linear regressions confirmed the influence of increased age and LH on the AMS score and IIEF5. Each additional year of increase in age caused an increase in the AMS score of 0.324 (p < .01) and a decrease in the IIEF5 score of 0.387 (p < .01). Likewise, a 1-unit increase in LH increases AMS score of 0.182 (p < .01) and reduces IIEF5 score of 0.231(p < .01). Other sex hormones did not show such significant correlations.

Considering the association between age and sex hormones or among sex hormones, variance inflation factors (VIFs) were estimated in the multiple linear regression and indicated multi-collinearity. TT, cFT, Bio-T and SHBG were correlated. Therefore, age, BMI, and LH were kept in the model, and then only one of other sex hormones was added into the model each time. All VIFs were less than 1.5. The correlation results were similar to the formers, and only age and LH were still the significant correlated factors with AMS or IIEF5.

Discussion

In the former paper, we have found cFT and Bio-T decreased, AMS and Erectile dysfunction got more severe with aging [12]. In the current paper, when AMS sub-scores were further analyzed, somatic and sexual sub-scores significantly changed with aging, but there was no impact on the psychological sub-score. The correlations were similar to other studies [2,13].

Pairwise correlations revealed that, in comparison to TT, Bio-T and cFT levels were more strongly associated with AMS or IIEF5. Considering the data did not completely fit normal distribution, we compared the Pearson’s correlation results and Spearman correlation (non-parametric statistical method) results. Both the correlation coefficients r and p values were similar. Therefore, we can assume that the non-normal distribution of variables have leveled off according to the law of large sample.

Liu’s and Antonio’s study also showed cFT levels were more reliable than TT levels for diagnosing LOH in middle-aged and elderly males [14,15]. Some studies reported testosterone treatment can significantly improve male ED and LUTS [16,17], however, data on the effect of exogenous testosterone on men are inconsistent [18]. In sub-analysis, no correlation was observed between any hormone level and the psychological section of the AMS, whereas the somatic and sexual sections were correlated with the levels of cFT, Bio-T, SHBG and LH. This was partially supported by results of Chen W’s [19]. However, when age was included into the multiple regression models, cFT, Bio-T and SHBG failed to yield any additional predicting information. Only LH was still the associated factor of AMS and IIEF5 scores. Partial correlation analysis supported these results, too. The results were supported by Kratzik’s study [20]. In Tajar’s paper, normal T and elevated LH were defined as compensated hypogonadism, compensated hypogonadism was associated predominantly with physical symptoms and was considered to a genuine clinical subgroup of LOH [21].

More and more research studies document the associations between AMS or erectile dysfunction and general unwellness [22,23]. García-Cruz E and Hamanoue N pointed out AMS values were not associated with testosterone values but rather were related to insulin resistance, particularly in subjects with moderately severe AMS values [24,25]. The Massachusetts Male Aging Study indicated that obesity is an independent risk factor for ED [26]. However, we did not find the relationship between BMI and AMS or IIEF-5. Waist circumference may be a better predictor of hypogonadism than BMI [27]. A greater severity of ED in men with LOH correlated with an increased waist circumference and a history of diabetes mellitus [28]. Interestingly, testosterone treatment not only improved TT and cFT levels, but also increased lean mass and decreased fat mass [29]. Corona G’s study found, testosterone treatment significantly improves erectile dysfunction in men with testosterone deficiency. However, the magnitude of the effect was lower in the presence of metabolic derangements, such as diabetes and obesity [30]. In other aspects, for example, men who work nonstandard shifts and have poor sleep quality are at increased risk for hypogonadal symptoms and sexual dysfunction [31]. Serum inorganic phosphorus had the highest correlation coefficient with IIEF-5 score in the middle-age [32].

The limitation of our research is that we did not considering too many other influence factors on reproductive health. We are following-up these subjects and will collect more information, such as demographic statistics, living style, quality of life index, and so on.

Our research population is ordinary middle-aged and elderly men, not including many patients with severe diseases. The decreases of reproductive health were mainly induced by aging and aging-related comorbidities. Based on our study results, for ordinary middle-aged and elderly men, to improve the reproductive health, attention should not be only focused on testosterone supplement. The next step in our research will pay more attention to the effect of healthy living style, such as nutritionally balanced diet, proper exercise, regular schedule, on general wellness, as well as reproductive health.

Disclosure statement

No conflict of interest.

Additional information

Funding

The project was supported by National Science and Technology Basic Work Program of China (2013FY110500).