Andropause symptoms and sickness absence in Japanese male workers: a prospective study

Abstract

Objective

The purpose of the study was to investigate the association between andropause symptoms and sickness absence in Japanese male workers over 2 years.

Methods

A baseline survey asking about andropause symptoms, along with blood sampling for testosterone level, was conducted in June 2009. A total of 418 men (mean age = 52.4 years, SD = 8.6) participated and were followed through 2011. Hazard ratios (HRs) and 95% confidence intervals (CIs) for sickness absence were calculated using Cox proportional hazard models.

Results

During the follow-up period, 31 of 35 participants who took sickness absences had physical illnesses. A higher andropause symptom score was associated with an increased risk of sickness absence. Testosterone deficiency (<350 ng/dL) was not associated with sickness absence. Among the subscales of andropause symptoms, the somatic symptom score was positively associated with sickness absence, whereas testosterone deficiency combined with high sexual symptoms was not associated with sickness absence. Results were similar when limited to sickness absence because of physical illness. No significant interaction between andropause symptoms and testosterone deficiency was found.

Conclusions

Non-specific andropause symptoms unrelated to testosterone deficiency were positively associated with sickness absence.

Keywords:

• Andropause symptoms
• testosterone deficiency
• sickness absence
• prospective study
• Japanese male workers

Introduction

Serum testosterone levels decline in men as they age. This age-related androgen deficiency, known as late-onset hypogonadism (LOH), contributes to various symptoms such as lack of energy, depression, decreased libido, and erectile difficulties [1]. Those symptoms are collectively termed “andropause” and are similar to menopausal symptoms in women [2]. LOH has been reported to correlate with obesity, poor metabolic status, metabolic syndrome, diabetes mellitus, and dyslipidemia [1,3]. According to a review article [4], cardiovascular mortality was related to testosterone levels, with lower testosterone levels associated with increased mortality. Furthermore, men with lower testosterone levels may be at a higher risk of developing atherosclerosis and coronary artery disease [4].

Sickness absences have a large adverse effect on workplace productivity [5]. Age is an important effect modifier of the association between individual and work-related risk factors of work disability [6]. For example, increasing age in men may be related to longer absences related to musculoskeletal disorders [7]. Additionally, testosterone deficiency related to aging may be predictive of sickness absence among male workers. However, few studies have investigated an association between low testosterone levels and sickness absences among male workers.

A poor psychosocial work environment affects mental well-being and increases the risk of sickness absence [8–12]. A systematic review examining the association between work and depression found that an adverse work environment, including a low ability to make decisions, high job stress, and bullying, increases depressive symptoms over time [13]. Adverse psychosocial work conditions may influence not only mental health but also physical health. For example, psychosocial stress has been studied in relation to the development of cardiovascular diseases [14], cancer [15], and hypertension [16].

According to studies that investigated associations between job stress, low testosterone levels, and andropause symptoms [17,18], job stress could modify the association between low testosterone levels and andropause symptoms. Specifically, highly demanding jobs intensified andropause symptoms for men with low testosterone levels, and longitudinally, when job demand levels increased, andropause symptoms increased regardless of the change in testosterone levels. Job stress factors could have an effect on andropause symptoms among male workers, and andropause symptoms could, in turn, be related to sickness absence. However, to our knowledge, there have been no studies investigating the associations between andropause symptoms and sickness absence.

The purpose of the present study was to investigate the association between andropause symptoms, including testosterone levels, and sickness absence in Japanese male workers. We investigated whether andropause symptoms and testosterone deficiency were associated with sickness absences, after adjusting for job stress levels. We also examined an interaction effect of andropause symptoms and testosterone deficiency in their association with sickness absences, as well as that of testosterone deficiency combined with high sexual symptoms based on recommendations for clinical diagnosis [19].

Materials and methods

Participants

Participants were recruited from employees of a mid-size shipbuilding company. An occupational physician distributed a printed description of the study and an informed consent form to employees 1 week before the annual health check-up in June 2009. Employees who had already taken the sick leave at the baseline survey were not targeted for study participation at the annual health check-up. Consenting employees completed questionnaires on andropause symptoms and had blood samples collected at the time of the health check-up. Questionnaires regarding job stress, depressive symptoms, and health-related behavior had already been distributed to all employees in December 2009. At the time of distribution of the questionnaires in December 2009, a result of blood sampling in June was fed back. Participation in the present study was on a voluntary basis. Among the 696 male employees, 418 individuals (response rate = 60.1%) agreed to participate in the study.

Sickness absence data from the date of study entry to 31 December 2011 were collected by the occupational physician. Participants were followed until the first episode of sickness absence. Information on the cause of the sickness absence was provided by the participants to the occupational physician who medically certified the absences. The length of the absence was recorded by labor-management personnel. Participants’ characteristics at baseline are shown in Table 1. Compared with employees who did not consent to participate in the study, study participants were older, had lower levels of education, were more likely to be managers, were more likely to take medication, and were less likely to have depressive symptoms.

Table 1. Participants’ characteristics at baseline.

		Analyzed male participants (n = 418)		Excluded male participants (n = 278)		p^†
Age	Mean and SD	52.4	8.6	36.7	14.4	<0.001
Education						<0.001
Higher education	%	24.2		32.4
Junior or high school	%	45.0		19.8
Missing	%	30.9		47.8
Job position						<0.001
Managers	%	38.5		7.9
Non-managers	%	29.2		44.6
Missing	%	31.6		47.5
Job demand						0.001
Low	%	24.2		18.7
Intermediate	%	20.8		14.7
High	%	23.0		19.1
Missing	%	32.1		47.5
Job control						<0.001
Low	%	22.0		18.7
Intermediate	%	28.0		18.3
High	%	17.2		14.7
Missing	%	32.8		48.2
Medication treatment^#						<0.001
With medication	%	14.4		4.0
Without medication	%	49.3		46.8
Missing	%	36.4		49.3
Depressive symptoms						<0.001
Depressed (≥16)	%	13.2		12.2
Nomal (≤15)	%	53.3		39.2
Missing	%	33.5		48.6
Testosterone (ng/mL)	Mean and SD	557.9	195.6
Testsoterone deficiency (<350 ng/dL)	%	12.0

^#Medication of antihypertensive, insulin/lower blood glucose, lower cholesterol.

^†Comparison between analyzed male participants and excluded participants based on a t-test for a continuous variable and a Chi-square test for a categorized variable.

SD: standard deviation.

Written informed consent was obtained from all study participants. Participants were not compensated for their participation in this study. The study protocol was approved by the ethics committee of Okayama Prefectural University (Okayama, Japan). Study methods have been reported in detail elsewhere [20].

Questionnaire

The Aging Males’ Symptoms (AMS) scale [20] was used to measure andropause symptoms in middle-aged males. The AMS scale is designed as a self-administered instrument used to assess symptoms of aging (independent from symptoms that are related to disease) and evaluate symptom severity over time. Translation into Japanese was done using international methodological recommendations [21]. The AMS scale comprises 17 questions, each of which has a response scale with 5 categories of severity ranging from 1 to 5 points. The total symptom score is calculated by adding the points for each question and there are also three sub-scores for psychological, somatic, and sexual symptoms.

To measure work environment characteristics, we used the job control (9 items) and job demand (5 items) scales from a Japanese version of the Job Content Questionnaire (JCQ) [22,23]. The Japanese version of the JCQ has been validated and tested for reliability [23]. The job control scale assesses the ability to make decisions, be creative on the job, and develop one’s ability. Job demand assesses the quantity of work, intellectual requirements, and time constraints of the job. Likert scale response options range from 1 (completely disagree) to 4 (completely agree). The score for each scale is calculated by summing the individual question scores according to JCQ guidelines [24]). The JCQ was measured in the baseline survey in December 2009. Job demand and job control scores were categorized as high, middle, or low based on the tertile of the scores (job demand: 29 and 33, and job control: 62 and 70).

Depressive symptoms were assessed using the Japanese translation of the Center for Epidemiologic Studies Depression Scale [25,26]. This scale contains 20 items on the frequency of experiencing specific depressive symptoms in 1 week, with a choice from 0 (not at all) to 3 (more than 5 days). A score of 16 or higher is generally used as the cut-off point indicating a high level of depression [25].

The questionnaire also included questions on the following background information: occupational position (management or general level), an education level (high school level and lower or college and higher), and medication (whether participants were taking antihypertensive, insulin injection, or antihyperglycemic drugs, or cholesterol-lowering drugs). The occupational position was categorized as “1” for management levels and “0” for general levels. Education levels were categorized as “1” for college and higher and “0” for high school level and lower. Medication treatment was categorized as “1” for answering one of the treatments (i.e. antihypertensive, insulin injection, antihyperglycemic drugs, or cholesterol-lowering drugs) provided.

Blood sampling

Blood samples were collected between 11:30 and 13:00 to comply with the various schedules of individual participants during the annual health check-up at this company. Serum testosterone levels were measured using a chemiluminescent immunoassay (CLIA; ng/dL), and the hormonal estimation was conducted by BML, Inc., Japan. Based on a preceding study [27], testosterone deficiency was defined as total testosterone below 12.1 nmol/L, which is below 350 ng/dL. Based on the raw data, testosterone was categorized into two categories: low (dummy variable as (1) and normal (0).

Statistical analyses

Comparisons of the study participants and non-participant employees were conducted using t-tests for continuous variables and Chi-square tests for categorical variables. Associations of andropause symptoms and testosterone deficiency with sickness absence were assessed using Cox proportional hazard models. Potential confounding factors of age, education, job position, medication treatment, and depressive symptoms were entered into the models. Based on tertile scores, the total and subscale scores for andropause symptoms were categorized as high, middle, and low (total score: 26 and 36; psychological symptoms: 6 and 10; somatic symptoms: 11 and 15; sexual symptoms: 8 and 12). The high and middle groups were investigated in relation to sickness absence and compared with the low group as the reference group. Based on the recommendations from the Lisbon 2018 International Consultation for Sexual Medicine [19], testosterone deficiency combined with high sexual symptoms was also investigated in relation to sickness absence. The total and subscale scores for andropause symptoms and testosterone after logarithmic transformation were entered into the models as continuous variables. For the analysis of the interaction between andropause symptoms and testosterone deficiency, a multiplicative variable was calculated and was entered into the model. Missing data on the potential confounding factors were categorized and taken into the models as dummy variables. All tests were two-tailed and p < 0.05 was considered statistically significant. Statistical analyses were performed with the computer program SPSS, version 25 (IBM, Armonk, NY, USA).

Results

Among the 35 workers who had sickness absences, 31 had a physical illness. Of these, 11 had injury of the extremities or shoulder, 11 had articular inflammation in the extremities, lower back, or neck, 5 had inflammation of the lung, liver, renal, or appendix, 4 had depression, 2 had a tumor, 1 had influenza, and 1 had a spasm. The person-years at risk were 1014.5.

A crude HR and an age-adjusted HR did not show significant associations between the total score of andropause symptoms and sickness absence because of all causes. Conversely, a fully-adjusted HR showed a statistically significant association between a high andropause symptom score and sickness absence (fully-adjusted HR = 2.50, 95% CI: 1.01–6.17). Moreover, when the total score of the andropause symptom scale was treated as a continuous variable, it was positively associated with an increased risk of sickness absence because of all causes (crude HR = 1.03, 95% CI: 1.00–1.06, age-adjusted HR = 1.03, 95% CI: 1.00–1.06, adjusted HR = 1.04, 95% CI: 1.01–1.07; Table 2). Testosterone deficiency was not significantly associated with sickness absence. When a continuous value of testosterone was entered into the model, testosterone levels were not significantly associated with sickness absence. Additionally, testosterone deficiency combined with high sexual symptoms was not significantly associated with ickness absence (Table 2).

Table 2. Associations between andropause symptoms, testosterone deficiency, and all-cause sickness absence.

	N	n	Person years at risk	Crude HR	95% CI		Age adjusted-HR	95% CI		Multivariable adjusted-HR^#	95% CI
Andropause symptoms total score
Low	144	8	360.2	1.00			1.00			1.00
Middle	145	13	350.5	1.68	0.69	4.04	1.71	0.71	4.13	1.76	0.70	4.39
High	129	14	303.8	2.10	0.88	5.00	2.15	0.90	5.15	2.50*	1.01	6.17
Continuous value of total score				1.03*	1.00	1.06	1.03*	1.00	1.06	1.04*	1.01	1.07
Testosterone
Normal (≥350ng/dL)	368	34	889.2	1.00			1.00			1.00
Deficiency (<350ng/dL)	50	1	125.4	0.21	0.03	1.51	0.21	0.03	1.53	0.20	0.03	1.50
Continuous value of testosterone^†				1.94	0.75	5.01	2.01	0.77	5.29	1.79	0.69	4.64
Testosterone deficiency and high sexual symptoms
Normal	407	10	998.9	1.00			1.00			1.00
Deficiency and high sexual symptoms	11	1	25.7	1.15	0.16	8.37	1.22	0.16	9.10	1.35	0.18	10.37

^#Age, education, job position, job demand, job control, medication treatment, and depressive symptoms were adjusted.

^†Logarithmic transformation was applied for a continuous value of testosterone.

*p < 0.05.

Among scores of the andropause symptom sub-scales, the somatic symptom sub-scale was positively associated with an increased risk of sickness absence because of all causes (fully-adjusted HR = 1.21, 95% CI: 1.07–1.37; Table 3). The group that was high in andropause symptoms tended to have a higher risk of sickness absence, even though this association was not statistically significant (fully-adjusted HR = 2.46, 95% CI: 0.80–7.60). The psychological and sexual symptom sub-scales were not associated with sickness absence (Table 3).

Table 3. Associations between subscales of andropause symptoms and all-cause sickness absence.

	N	n	Person years at risk	Multivariable adjusted-HR^#	95% CI
Somatic symptoms
Low	159	9	397.1	1.00
Middle	131	9	320.5	1.42	0.50	4.04
High	128	17	296.9	2.46	0.80	7.60
Continuous value of somatic symptoms				1.21*	1.07	1.37
Psychological symptoms
Low	160	11	391.8	1.00
Middle	139	11	341.4	0.78	0.30	2.06
High	119	13	281.3	0.84	0.26	2.74
Continuous value of psychological symptoms				0.91	0.77	1.07
Sexual symptoms
Low	156	11	386.9	1.00
Middle	142	10	344.5	1.28	0.50	3.27
High	120	14	283.2	1.44	0.48	4.28
Continuous value of sexual symptoms				0.98	0.88	1.10

^#Age, education, job position, job demand, job control, medication treatment, depressive symptoms, and testosterone were adjusted.

*p < 0.05.

When limited to the 31 absences because of physical illness, the results of the analysis based on continuous values were similar (total score: crude HR = 1.03, 95% CI: 1.00–1.06, age-adjusted HR = 1.03, 95% CI: 1.00–1.06, adjusted HR = 1.04, 95% CI: 1.00–1.07; somatic symptoms: crude HR = 1.21, 95% CI: 1.06–1.37, age-adjusted HR = 1.20, 95% CI: 1.06–1.36, adjusted HR = 1.24, 95% CI: 1.08–1.41). Individuals who were high in andropause symptoms showed a non-statistically significant increased risk of sickness absence because of physical illness (full-adjusted HR = 2.16, 95% CI: 0.84–5.51).

When a multiplicative variable of andropause symptoms and testosterone deficiency was entered into the model, there was no significant interaction effect on sickness absence.

Discussion

The main findings of our study were that andropause symptoms overall, as well as somatic symptoms, were positively associated with an increased risk of sickness absence in Japanese male workers. However, low testosterone levels were not predictive of sickness absence and there was no interaction effect between andropause symptoms and testosterone deficiency on risk of sickness absence. Testosterone deficiency combined with high sexual symptoms was not associated with the risk of sickness absence. The results were similar after adjusting for job stress factors.

The andropause symptoms reported by middle-aged men include insomnia, decreased libido, reduced sexual activity, decreased mineral bone density, and abdominal obesity [28], and these are similar to menopausal symptoms in middle-aged women. Men with testosterone deficiency tend to display a similar syndrome, specifically in relation to the sexual domain, according to the European Male Aging Study survey [29,30]. Based on recommendations for clinical diagnosis [19], andropause refers to the combination of low testosterone levels and clinical symptoms, especially sexual symptoms. The tool used to measure andropause symptoms in the present study, the AMS scale, comprises not only “sexual symptoms,” but also “somatic” and “psychological” symptoms [31]. The results of this study were that sexual symptoms were not associated with sickness absence and neither were low testosterone levels. Additionally, no association was found between testosterone deficiency combined with high sexual symptoms and sickness absence. However, somatic symptoms were positively associated with an increased risk of sickness absence, particularly absences because of physical illness. Among the items in the AMS somatic symptom sub-scale was a question on “complaints in joints or muscles (pain in lower back, joints, or legs).” A higher percentage of men with testosterone deficiency reported back and/or neck pain than those without testosterone deficiency [32]. Somatic symptoms related to aging such as joint and muscle pain could be reflected on the scores of the AMS scale; however, those somatic symptoms are not specific for either andropause or LOH.

Few men with low testosterone levels took sickness absences in the present study. This finding is consistent with several previous studies that found that the biochemical demonstration of testosterone deficiency is not sufficient to define the clinical condition of LOH, and the presence of specific symptoms is an essential requirement [33,34]. The symptoms considered most specific are sexual symptoms, including erectile problems and interest in sex [33]. A Japanese study reported that because total testosterone levels in Japanese men were not associated with advancing age, they could not be used to diagnose LOH [35]. A potential reason that testosterone deficiency is insufficient is that consensus is lacking about the level of serum testosterone that defines testosterone deficiency [36]. Several thresholds have been identified, including <300 ng/dL, <200 ng/dL, <337 ng/dL, between 231 and 346 ng/dL [3]. In the present study, low testosterone levels were defined as below 350 ng/dL, but only 50 men were in this category and only one of these men took a sickness absence. In addition to the difficulty of defining a cutoff for a low testosterone level, testosterone levels may vary over time because of factors such as circadian and seasonal variations, as well as changes because of aging. Large cross-sectional studies have consistently found a 1–2% annual decline in free or bioavailable testosterone, offset by a rise in SHBG, which results in a net decline in total testosterone [36]. The rate of decline is the same in healthy men as in those reporting chronic illness, obesity, alcoholism, prescription medication, or prostate problems [36]. On the basis of the recommendations for clinical diagnosis [19], free testosterone may be a useful indicator of androgen status. Furthermore, testosterone deficiency occurs following changes in luteinizing hormone levels as well as when serum testosterone is low [19]. However, data for free testosterone and luteinizing hormone levels were not available in the present study.

Levels of dehydroepiandrosterone (DHEA), one of the other androgens, may be predictive of sickness absence in men. Our previous study found that job stress levels were associated with serum DHEA sulfate (DHEA-s) and we noted that job stress may alter hormonal secretion, which, in turn, could be reflective of physical damage [20]. DHEA and DHEA-s, which are secreted by the adrenal cortex, have been found to protect against stress and have a regenerative role in response to stressors [37–39]. Concentrations of DHEA and DHEA-s show a progressive age-related decline in men as well as women [40,41]. DHEA is an inactive precursor of active sex steroids in peripheral target tissues [42] Thus, a decline in DHEA levels with aging could be linked to a reduction of testosterone as well as to somatic complaints, which in turn, could lead to more serious illness. In the present study, there was no significant association between DHEA-s levels and andropause symptom scores (data not shown). However, other hormonal changes associated with aging, along with perceived andropause symptoms, should be investigated in relation to sickness absence in future studies.

The present study was the first prospective study to find associations between andropause symptoms and sickness absences. However, our study had several limitations. Because the sample size of the baseline survey was small and not representative of Japanese working men, the results may not be generalizable. Moreover, individuals who took sickness absence because of mental health issues did not consent to participate in this study; the 30 month follow-up period may have been too short to include this group of men. Because multiple factors may underlie perceived symptoms and lead to sickness absences, other potential confounding factors should be considered for future studies.

In addition to the above potential limitations, the present study was conducted in Japan and its findings may not be generalizable across cultures. According to Daig et al. [43], the internal structure of the AMS across countries was sufficiently similar and the scale measures the same phenomenon. However, there may be cultural differences in andropause symptoms among individuals using this scale. Moreover, both the total and free testosterone levels in Japanese men were lower than those in the American men who participated in the Framingham Heart Study [35]. Those cultural and physiological differences should be taken into consideration when interpreting the results of this study.

In conclusion, this study found that only non-specific andropause symptoms, especially somatic symptoms, were positively associated with sickness absences. However, no such associations were found for low testosterone levels either alone or when combined with high sexual symptoms. A decline in testosterone has been linked with hyperlipidemia, hypertension, back and/or neck pain, and human immunodeficiency virus/acquired immunodeficiency syndrome [32]. Those symptoms could be perceived as somatic symptoms in the present study. It has been noted that the variety of non-specific symptoms that show associations with testosterone deficiency may make it difficult to establish a diagnosis of LOH [33]. The impacts of andropause symptoms and physiological aging on health status among middle-aged working men should therefore be investigated in further studies.

Author contributions

Kumi Hirokawa designed the study, analyzed the data, and wrote the manuscript.

Yoshihito Fujii participated in data collection and was in charge of dataset management.

Toshiyo Taniguchi participated in data collection and management of the study protocol.

Jiro Takaki participated in data analysis and manuscript writing.

Akizumi Tsutsumi participated in manuscript writing and revised the manuscript.

Acknowledgments

The authors thank Diane Civic, Ph.D., from Edanz Group (https://en-author-services.edanzgroup.com/ac) for editing a draft of this manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability

The data that support the findings of this study are available from the corresponding author, [KH], upon reasonable request.

Additional information

Funding

This study was supported by Japan Society for the Promotion of Science, KAKENHI [21700681].