Assessing symptoms of hypogonadism by self-administered questionnaire: qualitative findings in patients and controls

Abstract

Background. Current screening instruments for hypogonadism lack adequate specificity and diagnostic accuracy. A new self-administered questionnaire of hypogonadism symptoms is being developed to address this need. The process for questionnaire development and results from the first (qualitative) phase are presented.

Methods. Qualitative interviews were conducted based on a new conceptual model of hypogonadism and according to standards for questionnaire development. An item pool was generated from focus groups and in-depth interviews with two groups of hypogonadal patients, treated (N = 26) and untreated (N = 26), and age-equivalent controls (N = 28). Standardized scoring of the qualitative interviews was used to confirm conceptual domains in the model and to generate questionnaire items for further validation.

Results. Key domains identified in both patients and controls included: (a) physical function; (b) bodily signs and symptoms; (c) sexual function and libido; (d) sleep function; (e) mood and affective function; (f) memory and cognitive function. The final domain is distress or bother associated with hypogonadism symptoms. This domain was only relevant to the patient groups.

Conclusions. The first stage in the design of a new hypogonadism screener has been completed. Seven domains were identified and draft items were developed in each domain according to current standards of patient-reported outcomes.

Keywords:

• Hypogonadism
• screening instrument
• patient-reported outcome
• qualitative interviewing
• conceptual model

View correction statement:

Erratum

Introduction

Hypogonadism (HG) in men is characterized by a marked reduction or deficiency in the production of serum testosterone. Men with HG may present clinically with various signs or symptoms. These may include low libido, decreased muscle strength and physical function, loss of energy, depressed mood, and impaired cognition [1-3]. Although not generally considered signs of low testosterone, numerous comorbidities have been associated with low testosterone, including abdominal obesity [4],[5], markers of pre-diabetes (insulin resistance, impaired glucose tolerance, and metabolic syndrome) [6],[7], and type 2 diabetes [6-9]. Elevated risk of mortality in men with low testosterone levels has been reported in some [10-13] but not all studies [14-16].

Most elderly men have testosterone levels within the normal range, with prevalence estimates of ‘low’ (e.g. <300 ng/dl (10.4 nmol/l)) serum testosterone generally between 10% and 25%[17-19]. Prevalence estimates of HG or symptomatic androgen deficiency based on both symptom and testosterone data are generally <10%[20],[21]. According to data from the Hypogonadism in Males Study [22] and the Boston area community health survey [23], <15% of hypogonadal men report use of testosterone therapy. The relatively low rates of testosterone use may be due to concerns regarding the safety of testosterone. Alternatively, it may be that a large percentage of patients with HG are currently undetected. This could be due to: (i) lack of knowledge among physicians about the condition; (ii) overlap of hypogonadal symptoms with symptoms associated with other medical conditions (e.g. depression, sleep apnea); and (iii) absence of a reliable and accurate screening tool. Currently available HG screening questionnaires [24-26] are lacking in specificity, and with the exception of the Aging Male Symptoms (AMS) rating scale, were not developed or validated according to current guidelines. Despite the fact that current guidelines explicitly recommend against screening for androgen deficiency in the general population [1], specific populations in whom screening for androgen deficiency may be advisable include men with osteopenia or low bone mass, erectile dysfunction (ED), diminished libido, and type 2 diabetes [27],[28].

This report presents results of a qualitative study that was designed specifically to develop an instrument that could be used for identification or classification of men with HG. The purpose of our study was to obtain subjective patient perspectives on the disorder to test and refine our conceptual model, and to develop domains and questionnaire items for a new outcome measure or screening tool. Qualitative research of this type ensures that the subjective experiences and perspectives of patients are adequately represented, and is essential to the process of development of a new clinical or patient-reported outcome (PRO) measure [29-31].

Design and methods

Conceptual domains and specific items for the instrument were developed based on a ‘triangulation’ strategy, as recommended by recent FDA guidance for the development of PROs [29]. Three major sources of data were included: (i) Patient self-reports in structured qualitative interviews; (ii) Evidence-based review of the literature on signs and symptoms of HG and available questionnaire measures; and (iii) Clinical expert opinion.

Conceptual model development

A new conceptual model was developed for understanding the symptoms and subjective experience of patients with HG and for guiding the development of qualitative interviews (Figure 1). On the basis of a comprehensive literature review and advice from clinical experts, it was hypothesized that patients with low testosterone, regardless of its origin, experience a broad range of subjective effects on mood, quality of life, mental alertness, and subjective well-being, in addition to physical signs and symptoms. The model was used as a guide for the selection of themes and topics for the qualitative interviews, as well as for selecting individual items for the screener as described below.

Figure 1. Conceptual model of HG including potential causes of, domains affected by, and outcomes associated with low testosterone levels. The grey shaded area indicates the focus of the current investigation.

Subjects

Hypogonadal and eugonadal subjects (age range 21–74 years) were recruited from multiple sources for this study. These included physician providers, community-based services, health forums, and media advertisements. Three groups of subjects were enrolled: treated (N = 26) and untreated (N = 26) hypogonadal patients as well as eugonadal controls (N = 28). Patients and controls were recruited in cohorts of six to eight individuals per cohort, and recruitment was continued until saturation was achieved across all the groups.

Diagnosed hypogonadal patients (treated and untreated) were recruited from the practices of three physicians who are knowledgeable in the diagnosis and management of HG (located in Brookline, MA (urology); Peabody, MA (endocrinology); and Swansea, MA (primary care)). Criteria for diagnosis of HG included: presence of clinical symptoms characteristic of HG (as judged by the physician), and a low total testosterone level (as evidenced by the patient's most recent testosterone level). Untreated hypogonadal men had not received any form of androgen replacement in the past 3 months. In addition, all patients and controls met the following criteria: ability to speak and read English; cognitive competence (as judged by the physician); and absence of any speech or comprehension difficulties. Control subjects had normal testosterone levels based on an early morning blood sample and were excluded for the presence of any major medical or psychiatric disorder (HIV/AIDS, cancer, heart failure, major depression or sleep apnea) and for taking certain medications (GnRH agonists/antagonists and androgen replacement therapy).

The study protocol and informed consent form were approved by the New England Research Institutes, Inc. (NERI) Institutional Review Board (IRB) and IRBs at clinical sites. All participants provided written, informed consent prior to participation. All subjects who participated in a focus group or individual interview were paid $50 for their participation. In addition, control subjects received $50 for the blood draw.

Individual interviews and focus groups

This study used standardized qualitative research methods. The qualitative data were collected through focus groups and in-depth individual interviews, as described below. A saturation approach – which is the point at which no new information or themes are forthcoming from the new sampled subjects [32] – was used to determine when to discontinue data collection. Trained qualitative researchers conducted interviews until saturation was achieved. Qualitative interviews were transcribed and analyzed according to standardized methods of qualitative analysis.

Focus groups

Three focus groups were conducted during the course of the study – one for each of the study groups. Each focus group ranged in size from four to six men. Focus groups were led by a trained focus group moderator using a semi-structured discussion guide that was developed based on the conceptual model. All groups began with a statement of purpose and review of procedures, including confidentiality, topics to be covered, and the need for candid responses. Subjects with HG were asked to discuss their experiences around the six conceptual domains with specific reference to their medical condition. Treated subjects with HG were also asked to consider their experiences prior to and during treatment. Control subjects were invited to discuss their experiences in each of these domains, but without specific reference to HG.

All focus groups were conducted in English and audio-recorded for transcription and analysis. Focus groups lasted 90–120 min. Audio files of the recorded focus groups were transcribed and analysts reviewed the transcripts to confirm the salience of the six conceptual domains, the relevance of the interview questions, and to identify newly emerging domains.

In-depth individual interviews

Once the recruitment quota for each focus group was met, subjects were recruited for semi-structured individual interviews. Individual interviews for hypogonadal patients were conducted either at the physician office that recruited an individual into the study, at the NERI offices, or at the subject's home. Individual interviews for control subjects were held at the NERI offices.

Individual interviews were conducted by a trained NERI interviewer following a discussion guide developed to cover the same six conceptual domains. Interviews were continued until saturation was reached. All interviews were audio-recorded for subsequent transcription and analysis. Individual interviews lasted 45–90 min.

Qualitative analysis

Analyses used both inductive and deductive approaches [33]. Following a deductive approach, analysts reviewed transcripts of the focus groups and individual interviews, identifying text relevant to the six conceptual domains. In addition, they identified themes or ideas pertinent to each domain. Using an inductive approach consistent with grounded theory [34], the analysts reviewed the transcripts for newly emerging themes or concepts in the text [35],[36]. The resulting domains and themes were independently reviewed by two of the authors, one experienced in qualitative methods (SAM) and the other familiar with clinical aspects of HG (ABA).

Questionnaire data

The AMS Scale was used to characterize the symptomatic status of the study sample and to assess concurrent validity of the conceptual domains. The AMS Scale is a 17-item questionnaire which assesses common symptoms associated with male aging and has been used in previous studies of HG [24].

Results

Characteristics of the sample

A total of 80 subjects were included in the study. Fifty-two were patients with HG (treated (N = 26), untreated (N = 26)) and 28 served as controls. They represented a diversity of ethnic and socioeconomic backgrounds (Table I). The majority of subjects were married and living with a partner in a heterosexual relationship. The age range of the study subjects was 22–74 years. Treated patients' ages ranged between 37 and 69 years, untreated patients between 39 and 74 years, and controls between 22 and 71 years. Hypogonadal patients were older on average than controls (57 years vs. 44 years). Treated patients were observed on an average of 50 months after their diagnosis, whereas untreated patients were observed on an average of 19 months after their diagnosis. Most treated (21 of 25 or 84%) and untreated (25 of 26 or 96%) patients had testosterone (T) levels <400 ng/dl at diagnosis. As expected, mean current testosterone levels in the three groups were appreciably different (treated patients, 427 ng/dl; untreated patients, 258 ng/dl; and controls, 505 ng/dl). Symptoms present (as judged by the physician) among patients at diagnosis are shown in Table II. Low libido, fatigue, and ED were the most commonly reported symptoms (Table II).

Table I.  Descriptive characteristics of the three samples: treated patients, untreated patients, and controls

Characteristic	Treated patients (N = 26)	Untreated patients (N = 26)	Controls (N = 28)
Age
Range	37–69	39–74	22–71
Mean (SD)	57.0 (9.1)	57.2 (10.1)	43.9 (13.0)
Hispanic, N (%)	0 (0)	0 (0)	2 (7.1)
Race, N (%)
White	21 (87.8)	25 (96.2)	21 (75.0)
Black or African-American	5 (19.2)	1 (3.9)	2 (7.1)
Asian	0 (0)	0 (0)	3 (10.7)
Native Hawaiian or other Pacific Islander	0 (0)	0 (0)	2 (7.1)
Other	0 (0)	0 (0)	0 (0)
Months since HG diagnosis*
Range	0.7–142	0.1–82	–
Mean (SD)	50.4 (43.1)	18.7 (23.3)	–
Testosterone level (ng/dl) at diagnosis^†
Range	42–602	79–438	–
Mean (SD)	296 (125)	233 (76)	–
Category
<100 ng/dl	1 (4%)	1 (4%)
100 to <200 ng/dl	5 (20%)	8 (31%)
200 to <300 ng/dl	8 (32%)	13 (50%)
300 to <400 ng/dl	7 (28%)	3 (12%)
400 to <500 ng/dl	2 (8%)	1 (12%)
≥500 ng/dl	2 (8%)	0 (0%)
Current testosterone level (ng/dl)^‡
Range	107–1201	129–438	309–969
Mean (SD)	427 (286)	258 (75)	505 (154)

*Computed as number of months between date of testosterone measurement used for diagnosis and date of screening.

^†One treated patient was missing information on testosterone level at diagnosis.

^‡Current testosterone level among controls represents the value obtained at screening.

Table II.  Symptoms present among patients at diagnosis, based on physician report/medical records

Symptom at diagnosis	Treated patients (N = 26), N (%)	Untreated patients (N = 26), N (%)
Low libido
No	1 (3.8)	7 (28.0)
Yes	23 (88.5)	18 (72.0)
Not evaluated	2 (7.7)	0 (0)
Erectile dysfunction
No	2 (8.0)	1 (3.8)
Yes	22 (88.0)	25 (96.2)
Not evaluated	1 (4.0)	0 (0)
Decreased spontaneous erections
No	2 (8.0)	8 (30.8)
Yes	9 (36.0)	17 (65.4)
Not evaluated	14 (56.0)	1 (3.8)
Decreased intensity of orgasm
No	2 (8.3)	11 (44.0)
Yes	4 (16.7)	13 (52.0)
Not evaluated	18 (75.0)	1 (4.0)
Infertility
No	4 (16.7)	22 (84.6)
Yes	0 (0)	1 (3.8)
Not evaluated	20 (83.3)	3 (11.5)
Fatigue
No	1 (3.8)	1 (3.8)
Yes	14 (53.8)	23 (88.5)
Not evaluated	11 (42.3)	2 (7.7)
Muscle weakness
No	1 (3.8)	13 (50.0)
Yes	10 (38.5)	11 (42.3)
Not evaluated	15 (57.7)	2 (7.7)
Loss of body hair/decreased shaving frequency
No	4 (16.7)	18 (69.2)
Yes	0 (0)	5 (19.2)
Not evaluated	20 (83.3)	3 (11.5)
Inability to concentrate
No	2 (8.3)	12 (46.2)
Yes	3 (12.5)	11 (42.3)
Not evaluated	19 (79.2)	3 (11.5)
Sleep disturbance
No	4 (16.7)	12 (46.2)
Yes	0 (0)	13 (50.0)
Not evaluated	20 (83.3)	1 (3.8)
Osteoporosis or fracture
No	4 (16.7)	21 (80.8)
Yes	0 (0)	2 (7.7)
Not evaluated	20 (83.3)	3 (11.5)
Small testes
No	4 (16.7)	21 (80.8)
Yes	0 (0)	2 (7.7)
Not evaluated	20 (83.3)	3 (11.5)
Hot flushes (sweats)
No	4 (16.7)	21 (84.0)
Yes	0 (0)	1 (4.0)
Not evaluated	20 (83.3)	3 (12.0)
Depressed mood
No	2 (8.0)	7 (28.0)
Yes	7 (28.0)	15 (60.0)
Not evaluated	16 (64.0)	3 (12.0)
Mild anemia
No	3 (12.5)	16 (61.5)
Yes	1 (4.2)	7 (26.9)
Do not know/could not identify	20 (83.3)	3 (11.5)
Increased body fat
No	3 (12.5)	9 (34.6)
Yes	2 (8.3)	15 (57.7)
Not evaluated	19 (79.2)	2 (7.7)
Decreased seminal fluid
No	25 (96.2)	15 (57.7)
Yes	1 (3.8)	11 (42.3)
Other symptoms*
No	23 (92.0)	12 (80.0)
Yes	2 (8.0)	3 (20.0)

*Other symptoms included gynecomastia, Peyronies disease, difficulty achieving orgasm, and memory complaints.

AMS scores for the three groups are shown in Table III. Overall and across domains, treated patients had similar scores to controls. Likewise, overall and across domains, untreated patients had higher mean scores than either treated patients or control subjects. Significance tests were not performed due to the descriptive nature of these data. (Table III)

Table III.  Aging Males' Symptoms subscale and total scores in the three samples: treated patients, untreated patients, and controls

AMS	Treated patients (N = 26)	Untreated patients (N = 26)	Controls (N = 28)
Somatic subscale
N*	24	25	27
Range	7–20	9–31	7–20
Mean (SD)	11.9 (2.8)	18.6 (6.4)	12.2 (3.7)
Psychological subscale
N*	25	23	26
Range	5–16	6–25	5–23
Mean (SD)	8.0 (3.3)	11.3 (4.9)	9.0 (4.1)
Sexual subscale
N*	25	26	28
Range	5–19	8–20	5–20
Mean (SD)	9.7 (4.5)	14.2 (3.6)	8.5 (3.7)
Total score
N*	23	23	25
Range	18–55	23–70	17–57
Mean (SD)	29.8 (9.0)	44.4 (12.4)	29.6 (10.4)

*N represents total number with non-missing data.

Qualitative data

Subjects with HG reported a range of sexual, psychological, and interpersonal aspects of their disorder. Common themes within each response domain were identified across the groups, as described below. Most themes were repeated across focus groups and in-depth interviews. Some of these themes were apparent in interviews with control subjects also (particularly domains related to sleep, affective and cognitive function), but these were isolated responses and not consistent or characteristic of the control group as a whole.

Major themes and response domains

Physical function

The four most commonly experienced physical function problems among patients with HG included fatigue or lack of energy, decreased muscle mass and strength, decreased balance, and decreased coordination. The most common functional symptom experienced by men with HG was persistent fatigue, illustrated by the following quote:

‘Completely exhausted. Could stay in the bed around the clock. Would even put off urinating as long as I could rather than get up and off the bed to go and urinate, completely exhausted’.

Additionally, the following are examples of quotes from men who experienced loss of muscle mass and strength, and decreased balance/coordination:

‘Say you're carrying groceries and you pick them up, you can't hold them as long as you used to hold them … The strength goes out of me quicker’.

‘When I put my pants on, I need to stand on one foot, and I [can't] do that’.

Of note, all of the patients with HG experienced one or more of these functional symptoms in varying degrees of severity. Fatigue and loss of muscle strength were reported as more bothersome and distressing overall than decreased balance or coordination) which were generally less distressing. While complaints of muscle loss, fatigue, and decreased balance/coordination were noted by some of the eugonadal controls, these symptoms were relatively infrequent and reported as less bothersome by eugonadal controls.

Bodily signs and symptoms

At least one or more physical changes and bodily symptoms were described by the majority of treated and untreated patients with HG. The most frequently reported physical changes included: weight gain; height loss; dry, itchy eyes; hair loss; shrinking testes; and breast enlargement. Weight gain was the most commonly experienced physical change associated with HG, as evidenced by the following quote:

‘I kept insisting that my weight and my tiredness and everything else wasn't due to over-eating or over-drinking or lack of exercise. It was just the opposite. I was working out four days a week. I was running five miles. I was playing squash seven days a week. And I was in good shape, but I was getting heavier and heavier … So I said something's not right’.

In contrast, control patients rarely experienced any of the aforementioned bodily signs or physical symptoms, with exception of dry, itchy eyes, which was experienced almost equally by both HG and control respondents.

Sexual function

Treated and untreated men with HG reported experiencing a number of distressing symptoms related to sexual function, including loss of sexual desire, and ED. Many men complained specifically that they were no longer able to achieve spontaneous erections at the thought of sex, as in the following example:

‘Even when I was thinking about sex, it wasn't happening spontaneously … It was the uncertainty of whether or not it was going to happen … All of a sudden there's a problem and how do you explain that something happened? … And that's what I hate, because I always like being in control and knowing’.

A number of men complained of difficulty maintaining their erection, or that they could no longer achieve a firm erection. In one man's words ‘the hardness of an erection isn't the same; it dwindles’. Additionally, men overwhelmingly reported the absence of morning erections, though this was not usually distressing. Many untreated men reported feeling a lack of desire or urge to have sex, and they described feelings of ‘loss of manliness’ and low self esteem, as well as strained relationships associated with their dwindling desire:

‘I used to feel that I had an extremely active libido, and that went to a very low libido. So I pretty much didn't initiate any kind of sexual activity. And then even when my wife initiated it, I felt that I wasn't able to do my part in that aspect of our relationship’.

Some men noted that while they continued to have interest in sexual relations, the loss of erection had led to diminished libido. Concerns about sexual desire were expressed to varying degrees by all treated patients. Some men reported that their desire was the same (i.e. diminished) both pre- and post-treatment, while other men noted substantial improvements in desire and erectile function with the treatment:

‘If I have the (testosterone) shot, there's no reduction in the desire. If you don't have the shot, then you have no desire’.

[With the testosterone]‘I don't go soft. If I want to continue, I can continue’.

Sexual function was markedly affected by low testosterone among treated subjects and most men reported improvement in this domain coinciding with their regular testosterone treatment. Some older controls reported dwindling desire over the years, but this was not a major concern for them, and in many cases they attributed loss of desire to other issues in their life such as stress in a relationship. Likewise, a handful of older control subjects reported diminished erectile function, which they associated with lack of interest in a partner or with normal effects of aging.

Sleep function

Many men with HG suffered from ‘feeling tired all the time’ as well as somnolence. Many men described a sense of sleepiness that seemed disproportionate to the amount of sleep they were getting, or to any other activities in their life:

‘Just being tired all the time. It just doesn't seem right. Even coming off a vacation I felt tired, and I was off most of the holidays. A couple days I ended up just falling asleep in the chair in the mid afternoon. That could be just a chance of my body catching up, but it just didn't seem right to be that tired all the time’.

Men in both the treated and untreated hypogonadal groups complained of restless sleep (difficulty staying asleep) more often than they cited difficulty falling asleep. As one man put it:

‘Typically I don't have a hard time falling asleep. I have a hard time staying asleep, in the first hour or so. Typically, if I wake up within the first hour of falling asleep, I'm up for several hours. I can't get myself back to sleep’.

Several men reported waking up tired or not feeling ‘refreshed’ upon waking. Responses did not differ markedly between treated and untreated men with HG for this domain. Sleep problems were generally infrequent and not distressing among controls.

Mood and affective function

Men with treated and untreated HG experienced varying degrees of anxiety and depression related to their difficulties with sexual function. Further, the effects of low testosterone on sexual performance led men to feel ‘self-loathing,’ or as if they were ‘less of a man’. They equated the ‘performance’ of pleasing their partner with their sense of masculinity, and the sudden loss of control over their body caused hypogonadal patients to express diminished self-confidence and self-worth and lack of enjoyment of life:

‘Being a man is just being a man. Just, you know. Being alive … Being a man in the sense of … having a good time, keeping your partner happy. Just enjoying life. And that's one part that being a man that I'm not enjoying’.

Some treated men experienced dramatic changes in this domain with their testosterone treatment, noting a marked difference in their mood and sexual prowess following treatment. Other men claimed that as the day of their bi-weekly treatment appointment approached, they experienced declines in mood, energy, and/or sexual function. For example, one patient directly linked testosterone to his feelings of self-worth:

‘I attribute a lot of the depression and anxiety to lack of self-esteem, which comes with testosterone. For me, without a doubt, one of the biggest benefits [of testosterone treatment] I get is self-esteem, because there's more energy and feeling more muscular and feeling more masculine. And that goes away when I'm not on the testosterone. Just dives right off. So I'm not confident to walk into my work place and stand up to the guys that I work with or walk into a store at Christmastime and elbow my way through, get what I'm looking for … Which then turns into depression, because you go, well, [expletive], I'm 37 years old and I can't walk into JC Penny and pick up a toy for my kid. You know?’

In addition, patients reported increased feelings of impatience, loss of temper, and increased irritability. Some men linked such affective factors to other symptoms such as loss of sleep, fatigue, and frustration over relationship or sexual troubles. Many men also reported lack of motivation connected to low energy, and related loss of productivity, which adversely affected many parts of their lives. For example, one patient attributed the loss of his job to discontinuing testosterone treatment:

‘I basically lost a position at work because … I went off the testosterone and lost my motivation to succeed. I lost my energy to go the extra mile to get projects done and stand up for what I believed in … Performance reviews and stuff were low for the very first time in my career’.

Overall, the emotional lives of men were drastically affected by low testosterone, and the majority of men on treatment experienced marked improvements in their affect. Controls occasionally reported experiencing anxiety, sadness, or irritability, but rarely in combination with each other, and not to the extent expressed by patients with HG. Mood changes were also less distressing for controls.

Cognitive function

Many treated and untreated subjects with HG expressed concerns about declining cognitive function or mental performance, such as difficulty in remembering routes of travel, names, and recently read information, difficulty performing sequential tasks, and difficulty concentrating. The following is a representative quote from one subject:

‘I used to … read a book in two days and tell you everything about it. Can't do that anymore. I don't really want to read a book anymore, because I have to keep going back over and over it’.

When asked about the reason for their cognitive difficulties, not all treated and untreated hypogonadal patients attributed experiences of diminished concentration to low testosterone; some men claimed to have experienced concentration difficulties for much of their lives, whereas other men blamed old age or other conditions, such as attention deficit disorder. One treated subject noted that he ‘got sharper with the testosterone [shots],’ but for the most part, treated men expressed similar concerns around concentration as untreated men, and consistent improvements were not noted following treatment.

Men in the control group had similar experiences associated with memory loss and low concentration as the treated and untreated men with HG. Most men in the control group associated their memory and concentration difficulties with age or recounted that they had experienced such difficulties all their life.

Discussion

An in-depth qualitative study was performed with hypogonadal patients and controls to assess the clinical relevance of our conceptual model and to develop domains and items for a new PRO measure of HG. Current instruments for assessing HG lack adequate specificity for use in clinical practice or research [24-26]. Furthermore, with the exception of the AMS, none of the existing measures were developed in accordance with FDA guidance on development of PROs. Given the growing clinical need and inadequacy of current instruments, this study was the initial stage towards the development and validation of a clinically-useful, evidence-based instrument for assessing HG in men.

On the basis of the qualitative findings reported here, we observed notable differences between the subjective responses of patients with HG and controls across all domains, though not across all themes. In particular, subjective reports of fatigue, weight gain, muscle weakness, balance problems, hair growth, decreased concentration, sexual problems (ED, low libido), restless sleep, and low self-esteem were frequent and bothersome problems for many of the patients with HG. Subjective reports of difficulties in these domains were infrequent and much less bothersome for controls. Control subjects tended to be able to attribute symptoms to a separate condition or event (e.g. weight gain due to over-eating), whereas patients with HG often felt their symptoms were unexplainable or not due to any external factors.

Patients receiving treatment generally reported significant improvements in these areas of function following initiation of testosterone treatment. Concurrent validity of these subjective differences is supported by the pattern of results observed between the three groups in testosterone levels and responses to the AMS questionnaire.

As recommended by current guidelines [28],[29],[34],[35], a diverse group of treated and untreated patients and eugonadal controls were included in our study. Given the small sample size and convenience sampling, it is not possible to compare patients or controls in this study to population norms for eugonadal or hypogonadal men. Significance testing was not used to compare study groups in view of the limited sample and the absence of a priori statistical hypotheses or power estimates. Significance testing and comparison with larger population-based studies will be reported in our forthcoming quantitative validation study in a larger sample of hypogonadal patients and controls.

Findings from the qualitative interviews in our study support the conceptual model that was developed based on systematic literature review and expert panel recommendations. Subjective responses from treated and untreated patients confirmed the clinical relevance and comprehensiveness of the six domains identified in our conceptual model. The interviews also highlighted the subjective importance of certain symptoms (e.g. fatigue) in characterizing hypogonadal patients in comparison to eugonadal controls. Additionally, the interviews provided a rich source of verbatim quotes and self-statements to be used as a basis for drafting specific questionnaire items. This new questionnaire will be systematically evaluated in a controlled, quantitative validation study according to current standards for validation of PROs [30],[35]. It is anticipated that the use of a rigorous approach for the development of a new outcome measure or screening tool and by considering the patient's perspective on HG will improve the psychometric properties (e.g. specificity) of the resulting questionnaire.

Acknowledgements

This study was supported by Endo Pharmaceuticals. The authors gratefully acknowledge the contributions of the expert scientific advisory panel: Shalender Bhasin, MD; Andre T. Guay, MD, FACP, FACE; Martin Miner, MD; Abraham Morgentaler, MD, FACS; Richard Sadovsky, MD; Ridwan Shabsigh, MD, FACS; Ronald Swerdloff, MD. Raymond Rosen is a paid consultant – Boehringer-Ingelheim, Inc.; consulting fees – Johnson & Johnson, Ferring Pharmaceuticals and Eli Lilly & Co. Andre Guay is an advisory board member for Endo Pharmaceuticals. Abraham Morgentaler is a consultant, scientific advisory board – Endo Pharmaceuticals, and a recipient of research grant. Martin Miner is incharge of research support – GlaxoSmithKline, Endo Pharmaceuticals and consultant for GlaxoSmithKline, Endo Pharmaceuticals, Auxilium. Andre Araujo, Megan Connor, Emily Elstad and Sarah McGraw have no conflicts of interest to declare.