Development of and Recovery from Secondary Hypogonadism in Ageing Men: Prospective Results

1 2 1Development of and Recovery from Secondary Hypogonadism in Ageing Men: Prospective Results 2from the EMAS

3Giulia Rastrelli1, Emma L. Carter2, Tomas Ahern2, Joseph D. Finn2, Leen Antonio3, Terence W. O’Neill4, 4Gyorgy Bartfai5, Felipe F. Casanueva6, Gianni Forti1, Brian Keevil7, Mario Maggi1, Aleksander Giwercman8, 5Thang S. Han9, Ilpo T Huhtaniemi10, Krzysztof Kula11, Michael E. J. Lean12, Neil Pendleton13, Margus 6Punab14, Dirk Vanderschueren3, Frederick C. W. Wu2 and the EMAS study group.

71Sexual Medicine and Andrology Unit, Department of Experimental Clinical and Biomedical Sciences, 8University of Florence, Florence, Italy; 2Andrology Research Unit, Centre for Endocrinology & Diabetes, 9Institute of Human Development, Old St Mary’s Building, The University of Manchester, Manchester, UK; 103Department of Andrology and Endocrinology, Katholieke Universiteit Leuven, Leuven, Belgium, 4Arthritis 11Research UK Centre for Epidemiology, Institute of Inflammation and Repair, Faculty of Medical and Human 12Sciences, Manchester Academic Health Science Centre, University of Manchester and NIHR Manchester 13Musculoskeletal Biomedical Research Unit, Central Manchester NHS Foundation Trust, Manchester, UK; 145Department of Obstetrics, Gynaecology and Andrology, Albert Szent-György Medical University, Szeged, 15Hungary; 6Department of Medicine, Santiago de Compostela University, Complejo Hospitalario 16Universitario de Santiago (CHUS); CIBER de Fisiopatología Obesidad y Nutricion (CB06/03), Instituto 17Salud Carlos III; Santiago de Compostela, Spain; 7Department of Clinical Biochemistry, University Hospital 18of South Manchester, Manchester, UK, 8Reproductive Medicine Centre, Malmö University Hospital, 19University of Lund, Sweden; 9Department of Endocrinology, Ashford and St Peter’s National Health Service 20Trust, Surrey, UK; 10Department of Surgery and Cancer, Institute of Reproductive and Developmental 21Biology, Imperial College London, Hammersmith Campus, London, UK; 11Department of Andrology and 22Reproductive Endocrinology, Medical University of Łódź, Łódź, Poland; 12Department of Human Nutrition, 23University of Glasgow, Glasgow, UK; 13School of Community Based Medicine, The University of 24Manchester, Hope Hospital, Salford, UK; 14Andrology Unit, United Laboratories of Tartu University Clinics, 25Tartu, Estonia. 26

27Abbreviated title: Secondary Hypogonadism: Development and Recovery

29Key terms: hypogonadotropic hypogonadism, testosterone, androgen deficiency, obesity, co-morbidity

31Word count: 3472

32Number of figures and tables: 6

3 1 4 5 6 34Corresponding Author and person to whom reprint request should be addressed:

35Giulia Rastrelli, MD, PhD

36Department of Experimental Clinical and Biomedical Sciences

37University of Florence

38Viale Pieraccini,6, 50139, Florence, Italy

39Phone: +39 3334969190

40Fax: +39 0554271371

41e-mail: [email protected]

43Grants and fellowships supporting the writing of the paper: The European Male Aging Study is funded

44by the Commission of the European Communities Fifth Framework Program “Quality of Life and

45Management of Living Resources” Grant QLK6-CT-2001-00258 and facilitated by the Manchester

46Biomedical Research Centre and the NIHR Greater Manchester: Clinical Research Network. Additional

47support was also provided by Arthritis Research UK and the National Institute for Health Research and the

48Manchester Biomedical Research Centre.

50Disclosure statement: GR, ELC, TA, JDF, LA, TWO, GB, FFC, GF, BK, MM, AG, TSH, KK, MEJL, NP,

51MP, DV have nothing to declare. ITH has research contracts with Ferring Pharmaceuticals and Pantarhei

52Bioscience. ITH consults for Novartis. FCWW has consulted for Bayer-Schering, Eli Lilly, Repros

53Therapeutics Inc. and Besins Healthcare and also participated in advisory board meetings and

54lectured on their behalf. FCWW has received grant support (2010-2014) from Bayer Schering AG,

55PAREXEL International and Besins Healthcare.

7 2 8 9 10 61Abstract

62Context: Secondary hypogonadism is common in ageing men; its natural history and predisposing factors

63are unclear.

64Objectives: 1)To identify factors which predispose eugonadal men (T ≥10.5nmol/L) to develop biochemical

65secondary hypogonadism (T<10.5nmol/L, LH≤9.4U/L) and secondary hypogonadal men to recover to

66eugonadism. 2)To characterize clinical features associated with these transitions.

67Design: Prospective observational general population cohort survey.

68Setting: Clinical research centres.

69Participants: 3369 community-dwelling men aged 40-79 yr in eight European centres.

70Intervention: Observational follow-up of 4.3 years.

71Main Outcome Measure: Subjects were categorised according to change/no change in biochemical gonadal

72status during follow-up into persistent eugonadal (n=1909), incident secondary hypogonadal (n=140),

73persistent secondary hypogonadal (n=123) and recovered from secondary hypogonadism to eugonadism

74(n=96). Predictors and changes in clinical features associated with incident secondary hypogonadism and

75recovery from secondary hypogonadism were analysed by regression models.

76Results: The incidence of secondary hypogonadism was 155.9/10,000/year, while 43.8% of men with

77secondary hypogonadism recovered to eugonadism. Incident secondary hypogonadism was predicted by

78obesity [BMI≥30kg/m2: odds ratio (OR)=2.86(95% confidence interval 1.67;4.90);p<0.0001],weight gain

79[OR=1.79(1.15;2.80);p=0.011] and increased waist circumference [OR=1.73(1.07;2.81);p=0.026

80and 2.64(1.66;4.21);p<0.0001, for waist circumference 94-102 and ≥102 cm, respectively]. Incident

81secondary hypogonadal men experienced new/worsening sexual symptoms [low libido, erectile dysfunction

82and infrequent spontaneous erections]. Recovery from secondary hypogonadism was predicted by non-

83obesity [OR=2.28(1.21;4.31);p=0.011], weight loss [OR=2.24(1.04;4.85);p=0.042], normal waist

84circumference [OR=1.93(1.01;3.70)p=0.048], younger age [<60yr OR=2.32(1.12;4.82);p=0.024] and higher

85education [OR=2.11(1.05;4.26);p=0.037], but symptoms did not show significant improvement.

86Conclusion: Obesity-related metabolic and lifestyle factors predispose older men to develop secondary

87hypogonadism, which is frequently reversible with weight loss.

11 3 12 13 14 89Introduction

90After the third decade T decreases in men by 0.4-2% per year (1). Besides ageing, other risk factors

91contribute substantially to the T decline, particularly and irrespective of age, obesity (1,2). There is also

92evidence suggesting that low T can promote fat accumulation (3) suggesting a bi-directional relationship

93between obesity and low T. Longitudinal data from the European Male Ageing Study (EMAS) showed that

94weight gain was progressively associated with a decline in T levels without a concomitant change in LH (4),

95compatible with secondary hypogonadism (sHG). Furthermore, weight loss was proportionately associated

96with increases in T (4), suggesting that sHG is potentially reversible.

97sHG accounts for more than 50% of men with low T in the general population (5) and in patients with sexual

98dysfunction (6). To better understand the natural history and clinical significance of sHG, it is important to

99further investigate longitudinally the role of obesity, relative to other potential risk factors, in predicting the

100development of and recovery from sHG.

101Symptoms of androgen deficiency in the presence of low T provide the diagnostic cornerstone of the

102syndrome of hypogonadism (7). Late-onset hypogonadism (LOH) has been stringently defined by us as

103subnormal T associated with three sexual symptoms (5,8). However, the cross-sectional association between

104low T and symptoms was attenuated after adjusting for body mass index (BMI) and comorbidities (5),

105underlining the multi-causal origin of putative symptoms of hypogonadism in ageing men. Moreover,

106obesity, independent of T, is associated with sexual (9) and psychological symptoms (10) as well as impaired

107physical activity (11). Confirming the appearance of these symptoms with the development of biochemical

108hypogonadism and/or their resolution following recovery to eugonadism (EUG) would support their

109relevance as specific clinical markers of androgen deficiency, important in the diagnostic workup of men

110with low T.

112The aim of the study was to identify predictors of, and symptoms associated with incident sHG (isHG) and

113recovery from sHG (rsHG) in middle-aged and older men from the general population.

115Materials and methods

116Participants and study design

15 4 16 17 18 117The EMAS design and methods have been previously described (12,13). Briefly, an age-stratified sample of

1183,369 men aged 40–79 (mean±SD:60±11) years were recruited from population registers in eight European

119centres: Manchester (UK), Leuven (Belgium), Malmö (Sweden), Tartu (Estonia), Lodz (Poland), Szeged

120(Hungary), Florence (Italy) and Santiago de Compostela (Spain). Participants attended research clinics at

121baseline and 4.3 years later (range 3.0–5.7 years) for follow-up assessments (12,13). During this period, 193

122men died, 334 were lost to follow-up and 106 were institutionalised or became too frail. Ethical approval for

123the study was obtained according to institutional requirements in each centre. All participants provided

124written, informed consent. They completed questionnaires at both baseline and follow-up (12,13) about

125smoking, alcohol consumption and currently treated comorbidities (1). Anthropometric measurements,

126Reuben’s physical performance test (PPT) and psychomotor processing speed (Digit Symbol Substitution

127Test (DSST)) were performed according to standardised methods (12,13). Physical, sexual and psychological

128symptoms were determined from responses to the MOS 36-item Short-Form health survey (SF36), the

129EMAS Sexual Function Questionnaire and the Beck Depression Inventory (BDI), respectively.

130Hormone measurements

131Single, fasting morning (before 10:00 h) venous blood samples were obtained at baseline and follow-up. T

132was measured by liquid chromatography–tandem mass spectrometry (LC–MS/MS), with paired baseline and

133follow-up samples analysed simultaneously. LH, FSH and SHBG were measured by the E170 platform

134electrochemiluminescence immunoassay (Roche Diagnostics). Free T was calculated using the Vermeulen

135formula (14). Intra- and inter-assay coefficients of variation (CV) were: T 4.0 and 5.6%; SHBG 1.7 and

1363.2%, LH 1.9 and 3.0% and FSH 1.8 and 5.3%, respectively. The lower limit of total T measurement was

1370.17 nmol/l (0.05 ng/ml). Insulin was assayed using chemiluminescence (CVs: 3.9% and 5%). Biochemistry

138and haematology parameters were performed with standardized measurements, undertaken in hospital

139laboratories in each centre. Insulin resistance was calculated using the homoeostasis model assessment of

140insulin resistance [HOMA-IR=fasting insulin (U/ml)×fasting glucose (mmol/l)/22.5] (15).

141Gonadal status

142Participants with T≥10.5 nmol/l were defined as EUG, and sHG when T<10.5 nmol/l and LH<9.4 U/L (5).

143Subjects were further categorised by their change in gonadal status into (a) persistent EUG (pEUG) – EUG

19 5 20 21 22 144at baseline and follow-up, (b) incident sHG (isHG) – EUG at baseline and sHG at follow-up, (c) persistent

145sHG (psHG) – sHG at baseline and follow-up, and (d) rsHG – sHG at baseline and EUG at follow-up.

146Statistical analysis

147Baseline differences between isHG and pEUG and between rsHG and psHG in hormone levels,

148anthropometrics, biochemistry, symptoms, health and lifestyle measures were initially evaluated by

149Student’s t-test for continuous variables and Χ2 test for categorical variables.

150Multiple regression models, adjusted for centre as a random-effect, were used to account for the hierarchical

151study design (individuals nested within centre). The relationships between gonadal status and putative

152predictors were assessed using multilevel binary logistic regression models, where gonadal status was the

153outcome, with the pEUG or psHG group being the referent for the analyses of predictors of isHG or rsHG,

154respectively. Nine factors were included as fixed-effect predictors: age, smoking status (current smoker -

155yes/no), alcohol intake (alcohol consumption for ≥5 days per week vs. less), education level [low

156(compulsory education only), medium (non-compulsory education below university level) or high (university

157education)], Physical Activity Scale for the Elderly ((PASE) score ≤78 vs. >78), Chronic widespread pain

158(yes/no), marital status (no partner, having a partner but not living together or having a partner and living

159together), comorbidity (presence/absence of at least one self-reported disorder), BMI (<25 kg/m 2, 25–29.9

160kg/m2, and ≥30 kg/m2) and waist circumference (WC) (<94 cm, 94-101.9 cm, and ≥102 cm).

161The relationship between gonadal status and clinical features of HG was investigated using binary logistic

162regression models with symptoms, dichotomized as stable or new/worsened, as the outcome, when assessing

163outcomes of isHG, and as stable or resolved/improved, when assessing outcomes of rsHG. A symptom was

164defined as new when absent at baseline and present at follow-up and worsened when present at baseline but

165with a lower severity grading than at follow-up. A symptom was defined as resolved when present at

166baseline and absent at follow-up and improved when present at follow-up with a lower severity grading than

167at baseline.

168Differences in clinical characteristics between isHG and pEUG or rsHG and psHG at baseline and changes

169over time were investigated by multiple logistic regression models adjusted for age (as a continuous

170variable), centre, baseline BMI (as a continuous variable), presence of at least one comorbidity at baseline,

171smoking, and alcohol intake.

23 6 24 25 26 172Linear regression models were used to evaluate the association between isHG or rsHG and baseline levels or

173percentage change of metabolic and hematologic parameters, expressed as continuous variables.

174For isHG predictor analysis, weight gain was defined as ≥5% increase of the baseline value. In the rsHG

175predictor analysis, weight loss was defined as ≥5% decrease of the baseline value.

176Results from logistic regression models are presented as odds ratios (OR) and 95% confidence interval (CI)

177and from linear regression models as β coefficients with 95% CI. All statistical analyses were conducted

178using SPSS for Windows 20.1 (IBM, Armonk, NY, USA).

180Results

181Natural history of secondary hypogonadism (sHG)

182Of the 3,369 men that participated in EMAS (Figure 1): 193 were excluded at baseline because of medical

183conditions (known pituitary-testicular diseases and medications affecting T); 165 died during the follow-up

184period; 93 were institutionalised or too frail to attend for follow-up and 314 were lost to follow-up for other

185reasons. Another 162 were excluded at follow-up because of medical conditions/medications, and 121 were

186excluded due to missing total T and/or LH levels at baseline and/or follow-up. Among the baseline attendees,

187the prevalence of sHG was 10.0% (n=318). In this sHG group, mortality and lost to follow-up rates were 8.2

188and 11.3%, compared to the entire cohort’s rates of 5.2% and 9.9%, respectively. In the analytical sample of

1892,268 men, 1,909 were pEUG, 140 isHG, 123 psHG and 96 rsHG (Figure 1). The prevalence of sHG at

190follow-up was 11.0% and the incidence of sHG from EUG was 6.28% in 4.3 years or 155.9 per 10,000 per

191year or 1.6% per annum. The recovery rate from sHG to EUG was 30.2% (96/318) or 42.9% (96/224

192excluding 94 subjects not attending the second assessment) in 4.3 years.

193Compared with the analytical sample, men who died, were institutionalized or were too frail to attend were

194older, had lower free T, higher SHBG and gonadotropins, but not significantly different total T levels and

195reported more diabetes mellitus, cardiovascular diseases and worse physical performance (Supplementary

196table 1). Conversely, subjects who were lost to follow-up for other reasons were similar to the analytical

197group, except for a higher prevalence of smoking and metabolic syndrome (MetS) and lower psychomotor

198processing speed.

199 Characteristics of iSHG

27 7 28 29 30 200Lower mean total T, free T, SHBG, LH and FSH levels were already apparent in iSHG men compared to

201pEUG, at baseline; these differences were replicated and amplified at follow-up (Table 1). isHG men were

202overweight/obese at baseline, with higher prevalence of comorbidities and MetS; these factors increased

203further at follow-up. isHG men also showed differences in baseline and follow-up metabolic profiles with

204lower high density lipoprotein (HDL), higher triglycerides, glucose and HOMA-IR.

205Multiple logistic regression modelling identified obesity [OR=2.86 (1.67; 4.90),p<0.0001], as an

206independent predictor of isHG (Figure 2A). Substituting WC for BMI categories showed that increased WC

207was a predictor of isHG [OR=1.73(1.07;2.81);p=0.026 and 2.64(1.66;4.21);p<0.0001for WC 94-102 cm and

208WC>102 cm respectively] and that older men (≥70yr) demonstrated a significantly lower predisposition to

209develop sHG compared to younger subjects [OR=0.51(0.28;0.96),p=0.035] (Figure 2C). Weight gain of ≥5%

210was also a risk factor for isHG [OR=1.79(1.15; 2.80),p=0.011]. No significant interaction effect was found

211between the variables in the model.

212Prevalence of symptoms at baseline and follow-up, as well as new/worsening symptoms during follow-up

213are shown in Table 1. Men with isHG, compared with pEUG, did not show any difference in prevalence of

214symptoms at baseline before or after adjustment (Table 2, Figure 3A). isHG was associated with a higher

215prevalence of decreased libido at follow-up and, although just missing statistical significance, the prevalence

216of erectile dysfunction (ED) and infrequent morning erections were higher than pEUG at follow-up (Table 2

217and Figure 3B). isHG, however, was significantly associated with the development or worsening of all three

218sexual symptoms and one physical symptom (impaired vigorous activity) (Table 2). Incremental adjustments

219for potential confounders (Table 2) showed that isHG maintained its association only with new/worsening

220sexual symptoms (Figure 3C).

221Compared with pEUG, isHG subjects had significantly higher total cholesterol at baseline (Table 3). In

222addition, isHG was associated with a significant increase in HDL-cholesterol (Table 3). The apparent

223relationship between isHG with increased HDL-cholesterol became insignificant (β=0.44 [-0.08;

2240.95],p=0.096) after adjusting for those starting lipid-lowering medications during follow-up. iSHG men did

225not differ from pEUG at baseline, but they showed a significant decrease in feeling of physical well-being

226(SF36-physical component score), and deterioration in PPT rating (Table 3).

31 8 32 33 34 228Characteristics of rsHG

229At baseline, rsHG subjects had higher SHBG and total T levels as compared with psHG (Table 1). rsHG men

230had a more favourable baseline metabolic profile, with lower triglycerides and a lower prevalence of MetS

231than psHG. They also showed lower weight, BMI and WC. rsHG was independently predicted by being

232younger and non-obese BMI [OR=2.32(1.12;4.82);p=0.024 and OR=2.28(1.21;4.31);p=0.011; respectively]

233(Figure 2B). Normal WC was a significant predictor for rsHG [OR=1.93(1.01;3.70) p=0.048] (Figure 2D). A

234higher level of education predicted rsHG [OR=2.11 (1.05;4.26);p=0.037] (Figure 2B and 2D). Weight

235reduction (≥5%) during follow-up, when substituted for BMI, was also a predictor for rsHG

236[OR=2.24(1.04;4.85);p=0.042]. No interaction effect was found between the variables in the model.

237Prevalences of symptoms at baseline and follow-up as well as recovery/improvement of symptoms during

238follow-up in psHG and rsHG are shown in Table 1. No significant association was found between rsHG with

239either baseline or follow-up prevalence or recovery/improvement of sexual, physical or psychological

240symptoms (Figure 3D). rsHG, compared with psHG, had significantly lower total cholesterol and

241triglycerides at baseline and a significant decrease in insulin during follow-up (Table 3), whereas no

242significant difference was found for physical and psychological function.

244Discussion

245The longitudinal data in this observational cohort of older men from the general population highlight the role

246of obesity and weight gain as the most important predictors for developing sHG, showing for the first time

247that isHG is associated with appearance or worsening of sexual symptoms only. Another new finding is that

248sHG is potentially reversible in a substantial proportion of men and that recovery is predicted by non-obesity,

249weight loss, younger age and higher education.

250Predictors of isHG

251As previously shown by cross-sectional data from EMAS, BMI is an important correlate of sHG (5). In this

252longitudinal evaluation, we are able to confirm that BMI predicts isHG. Our finding of obesity as a

253determinant for T decline, independent of LH, confirms earlier longitudinal results (16,17). The association

254between obesity and HG is complex and poorly understood. Obesity can induce peripheral and central

255insulin resistance (18), pro-inflammatory cytokine production (TNFα and IL-6) from adipocytes (19) and

35 9 36 37 38 256central nervous system endocannabinoid release (20) that can induce down-regulation of hypothalamic

257function. In addition, adipocytokines such as leptin and adiponectin have been shown to modulate GnRH and

258gonadotropin secretion (21) and to influence testicular T production (22). Conversely, the suggested role for

259an excess of estrogens, due to an increased aromatase activity in obese subjects, in reducing the GnRH-

260gonadotropin secretion has not been confirmed in obese diabetic (23) and non-diabetic men (5), their

261estradiol levels being lower, rather than higher and correlated with T levels but not with BMI.

262Our results also demonstrated a lower probability that older men will develop sHG. Cross-sectional data

263from EMAS have shown that ageing is associated with elevated gonadotropins and primary HG (1,5).

264Disorders causing a derangement of hypothalamic-pituitary axis superimposed on a background of high

265gonadotropin (as in older men) are less likely to suppress LH to values below the threshold used to define

266sHG. Men attending both phases of the study are healthier as compared with those who did not (see

267Supplementary table 1). It is therefore conceivable that among those attending for follow-up, a survival bias

268may have selected healthier older men who are less prone to develop sHG (24).

270Developing sHG: clinical features

271In this study, development of sHG was associated, independently of BMI, comorbidities and lifestyle, with

272new or worsening sexual symptoms but not physical or psychological ones. This is consistent with our

273previous cross-sectional analyses (8) indicating that these three sexual symptoms are the most specific

274subjective features associated with low T. In addition to confirming the substantial baseline prevalence

275(22.7-48.5%) of sexual symptoms, occurrence of new or worsening symptoms in the pEUG men is not

276negligible (18%), albeit at a lower rate than in isHG men (25.6–28.1%), thus accounting for the weaker

277association between isHG and increased prevalence of sexual symptoms at follow-up. This emphasises the

278importance of the co-occurrence of all three sexual symptoms for increasing the probability of a robust

279syndromic clustering with low T (8) which serves to provide the operational definition of symptomatic HG

280or LOH. The longitudinal results, showing a specific association between the development of sHG with the

281appearance or worsening of all three sexual symptoms prospectively, strongly endorse this approach.

282In the cross-sectional analyses, physical or psychological symptoms did not cluster together with low T, but,

283one of the physical symptoms, decreased vigour, did show an inverse correlation with T levels (8).

39 10 40 41 42 284Interestingly, in the longitudinal study, decreased vigorous activity is the only non-sexual symptom whose

285new occurrence or worsening was significantly associated with isHG, until smoking and alcohol habits were

286included as covariates. This emphasises the non-specific nature of non-sexual symptoms in sHG. However,

287when considering the objective evaluation of physical performance and psychomotor processing speed (PPT

288rating), a significantly greater reduction of these functions was found in isHG men, even though this was not

289translated into a subjective perception of deterioration.

291Predictors of rsHG

292The present results show that sHG is frequently (30.2%) reversible. Reversion to eugonadism was predicted

293by lower BMI or WC, weight loss, younger age and higher education. The weight associations are

294compatible with recent evidence from general population showing that weight loss is associated with a

295proportionate increase in T levels (4,25). Thus, a >5% weight loss was associated with a significant increase

296of T levels with the maximum effect in those men who lost ≥15% of weight when LH as well T increased

297contemporaneously (4). Supervised dieting/exercise or bariatric surgery in obese men has been shown to

298increase T and LH levels proportionally to the amount of weight loss (26). Interestingly, T increase attained

299in treated subjects was higher in younger patients (26). We found, in addition to lower weight, that younger

300age was an independent predictor of rsHG suggesting that the potential for weight loss to improve

301hypothalamic-pituitary function may be eroded by adverse effects of ageing. A novel finding is that a higher

302level of education was an independent predictor of rsHG. Education is a socio-economic status surrogate and

303an important determinant of health inequality. Higher education is a protective factor for cardiovascular

304diseases and all-cause mortality (27,28). Changes in behaviour in response to health education are most

305beneficial in higher socio-economic groups (27-29) which may translate into improved metabolic status and

306avoidance of obesity. Taken together, these results strongly suggest that sHG or functional hypothalamic

307suppression associated with obesity, is potentially reversible with weight loss and amenable to lifestyle-

308influenced metabolic modifications especially in younger educated men. Weight loss, especially in older

309men, can be unintentional and due to comorbidities, cancer cachexia or other wasting diseases. However,

310excluding men with incident cancer or heart diseases in our analyses did not influence the relationships

311between BMI or weight loss with rsHG (data not shown), suggesting that the observed weight loss in those

43 11 44 45 46 312obese middle-aged men (mean age 56 yr) with rsHG may well have been intentional and that the

313concomitant rise in T is part of the overall, positive health outcome.

315 Recovering from sHG: clinical features

316In this study, the anticipated association between rsHG and resolution/improvement of symptoms was not

317found. There are a number of possible explanations. The relatively small sample size may not provide

318sufficient statistical power to detect small differences between psHG (n=123) and rsHG (n=96), with a

319minority only (2-11 men) showing symptomatic improvements. Both groups are subjected to uncontrolled

320multiple and multi-directional influences during follow-up in this non-interventional observational study.

321The increase in T documented at one time point at follow-up may not represent a sufficiently sustained

322improvement to transmute into subjective symptomatic recovery. Randomized controlled trials of T

323replacement therapy demonstrating improvements in sexual function (30), physical strength (31) and

324depressive symptoms (32) usually show a 2-fold increase of T levels into the mid-normal physiological range

325during treatment, for a minimum of 24 weeks. The spontaneous rise in T levels observed in our rsHG

326subjects is relatively small by comparison (9.2 nmol/L at baseline increasing to 13.3 nmol/L at follow-up or a

32745% increment) and may not be sufficient to drive improvements or resolution of sexual symptoms. Given

328the high prevalence of sexual symptoms irrespective of the concentration of circulating T, a further

329explanation could be that important non-hormonal (e.g. psychosocial and relational) factors contribute to the

330persistence of symptoms, even after restoration of normal T levels .

332Strengths and limitations

333The strengths of this study include a large unbiased sample from the general population, the prospective

334design providing non-interventional longitudinal data to investigate the natural history and predictors of

335hypogonadism and the standardised instruments applied across centres and between the phases of the study.

336T levels were measured by LC–MS/MS with paired baseline and follow-up samples analysed

337simultaneously. Limitations in EMAS have been described previously (13). The study interval of 4.3 years

338may be relatively short for capturing the more subtle changes in signs and symptoms of HG in an

339observational study. A single LH measurement only was available to categorise hypogonadism, but its co-

47 12 48 49 50 340linearity with FSH would have minimised any potential misclassification. Also for T, a single measurement

341only was available at each time point. However, T is a stable analyte (33) and single measurements of T on

342morning samples, can provide representative and reliable data in large epidemiological studies such as the

343EMAS.

345Conclusions

346Our longitudinal data showed that overweight, obesity or weight gain predisposed older men to develop

347sHG. Older men were at lower risk. Development of sHG was associated with the appearance of new or

348worsening sexual, but not physical and psychological symptoms. sHG frequently remits spontaneously; this

349is predicted by lower BMI or WC or weight loss, as well as younger age and higher education. Biochemical

350reversal of sHG to EUG however was not accompanied by a significant symptomatic improvement. Further

351studies are indicated to corroborate and extend the present results.

51 13 52 53 54 352ACKNOWLEDGEMENTS

353The European Male Aging Study is funded by the Commission of the European Communities Fifth

354Framework Program “Quality of Life and Management of Living Resources” Grant QLK6-CT-2001-00258

355and facilitated by the Manchester Biomedical Research Centre and the NIHR Greater Manchester: Clinical

356Research Network. Additional support was also provided by Arthritis Research UK and the National Institute

357for Health Research and the Manchester Biomedical Research Centre. The Principal Investigator of EMAS is

358Professor Frederick Wu, MD; Andrology Research Unit, University of Manchester, Manchester, UK. The

359authors wish to thank the men who participated in the eight countries, the research/nursing staff in the eight

360centres: C Pott (Manchester), E Wouters (Leuven), M Nilsson (Malmö), M del Mar Fernandez (Santiago de

361Compostela), M Jedrzejowska (Łódź), H-M Tabo (Tartu), A Heredi (Szeged) for their data collection, and C

362Moseley (Manchester) for data entry and project co-ordination. Dr. D Vanderschueren is a senior clinical

363investigator supported by the Clinical Research Fund of the University Hospitals Leuven, Belgium.

55 14 56 57 58 380References

3811) Wu FC, Tajar A, Pye SR, Silman AJ, Finn JD, O’Neill TW, Bartfai G, Casanueva F, Forti G,

382Giwercman A, Huhtaniemi IT, Kula K, Punab M, Boonen S, Vanderschueren D. Hypothalamic-

383pituitary-testicular axis disruptions in older men are differentially linked to age and modifiable risk factors:

384the European Male Aging Study. J Clin Endocrinol Metab. 2008; 93:2737–2745

3852) Travison TG, Araujo AB, Kupelian V, O’Donnell AB, McKinlay JB. The relative contributions of

386aging, health, and lifestyle factors to serum testosterone decline in men. J Clin Endocrinol Metab.

3872007;92:549–555

3883) Saad F, Aversa A, Isidori AM, Gooren LJ. Testosterone as potential effective therapy in treatment of

389obesity in men with testosterone deficiency: a review. Curr Diabetes Rev 2012; 8:131-143

3904) Camacho EM, Huhtaniemi IT, O'Neill TW, Finn JD, Pye SR, Lee DM, Tajar A, Bartfai G, Boonen

391S, Casanueva FF, Forti G, Giwercman A, Han TS, Kula K, Keevil B, Lean ME, Pendleton N, Punab

392M, Vanderschueren D, Wu FC; EMAS Group. Age-associated changes in hypothalamic-pituitary-

393testicular function in middle-aged and older men are modified by weight change and lifestyle factors:

394longitudinal results from the European Male Ageing Study. Eur J Endocrinol. 2013;168:445-455

3955) Tajar A, Forti G, O'Neill TW, Lee DM, Silman AJ, Finn JD, Bartfai G, Boonen S, Casanueva FF,

396Giwercman A, Han TS, Kula K, Labrie F, Lean ME, Pendleton N, Punab M, Vanderschueren D,

397Huhtaniemi IT, Wu FC; EMAS Group. Characteristics of secondary, primary, and compensated

398hypogonadism in aging men: evidence from the European Male Ageing Study. J Clin Endocrinol Metab.

3992010;95:1810-1818

4006) Corona G, Maseroli E, Rastrelli G, Sforza A, Forti G, Mannucci E, Maggi M. Characteristics of

401compensated hypogonadism in patients with sexual dysfunction. J Sex Med. 2014;11:1823-1834

4027) Bhasin S, Cunningham GR, Hayes FJ, Matsumoto AM, Snyder PJ, Swerdloff RS, Montori VM;

403Task Force, Endocrine Society. Testosterone therapy in men with androgen deficiency syndromes: An

404Endocrine Society clinical practice guideline. J Clin Endocrinol Metab 2010;95:2536–2559

4058) Wu FC, Tajar A, Beynon JM, Pye SR, Silman AJ, Finn JD, O'Neill TW, Bartfai G, Casanueva FF,

406Forti G, Giwercman A, Han TS, Kula K, Lean ME, Pendleton N, Punab M, Boonen S,

59 15 60 61 62 407Vanderschueren D, Labrie F, Huhtaniemi IT; EMAS Group. Identification of late-onset hypogonadism

408in middle-aged and elderly men. N Engl J Med. 2010;363:123-135

4099) Han TS, Tajar A, O'Neill TW, Jiang M, Bartfai G, Boonen S, Casanueva F, Finn JD, Forti G,

410Giwercman A, Huhtaniemi IT, Kula K, Pendleton N, Punab M, Silman AJ, Vanderschueren D, Lean

411ME, Wu FC; EMAS group. Impaired quality of life and sexual function in overweight and obese men: the

412European Male Ageing Study. Eur J Endocrinol 2011;164:1003‐1011

41310) Luppino FS, de Wit LM, Bouvy PF, Stijnen T, Cuijpers P, Penninx BW, Zitman FG. Overweight,

414Obesity, and Depression: A Systematic Review and Meta-analysis of Longitudinal Studies. Arch Gen

415Psychiatry 2010;67:220-229

41611) Riebe D, Blissmer BJ, Greaney ML, Garber CE, Lees FD, Clark PG. The relationship between

417obesity, physical activity, and physical function in older adults. J Aging Health. 2009;21:1159-1178

41812) Lee DM, O'Neill TW, Pye SR, Silman AJ, Finn JD, Pendleton N, Tajar A, Bartfai G, Casanueva F,

419Forti G, Giwercman A, Huhtaniemi IT, Kula K, Punab M, Boonen S, Vanderschueren D, Wu FC,

420EMAS study group 2009 The European Male Ageing Study (EMAS): design, methods and recruitment. Int

421J Androl 32:11-24

42213) Lee DM, Pye SR, Tajar A, O'Neill TW, Finn JD, Boonen S, Bartfai G, Casanueva FF, Forti G,

423Giwercman A, Han TS, Huhtaniemi IT, Kula K, Lean ME, Pendleton N, Punab M, Silman AJ,

424Vanderschueren D, Wu FC, EMAS study group. Cohort profile: the European Male Ageing Study. Int J

425Epidemiol. 2013;42:391-401

42614) Vermeulen A, Stoica T, Verdonck L. The apparent free testosterone concentration, an index of

427androgenicity. J Clin Endocrinol Metab 1971;33:759-767

63 16 64 65 66 42815) Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model

429assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations i

430n man. Diabetologia 1985;28:412-419

43116) Derby CA, Zilber S, Brambilla D, Morales KH, McKinlay JB. Body mass index, waist circumference

432and waist to hip ratio and change in sex steroid hormones: The Massachusetts male ageing study. Clin

433Endocrinol (Oxf). 2006;65:125–131

43417) Mohr BA, Bhasin S, Link CL, O'Donnell AB, McKinlay JB. The effect of changes in adiposity on

435testosterone levels in older men: longitudinal results from the Massachusetts Male Aging Study. Eur J

436Endocrinol. 2006;155:443-452.

43718) Porte Jr D, Baskin DG, Schwartz MW. Insulin signaling in the central nervous system: a critical role

438in metabolic homeostasis and disease from C. elegans to humans. Diabetes 2005;54:1264–1276

43919) Pittas AG, Joseph NA, Greenberg AS. Adipocytokines and insulin resistance. J Clin Endocrinol

440Metab. 2004;89:447–452

44120) Pagotto U, Marsicano G, Cota D, Lutz B, Pasquali R. The emerging role of the endocannabinoid

442system in endocrine regulation and energy balance. Endocr Rev. 2006;27:73–100

44321) George JT, Millar RP, Anderson RA. Hypothesis: kisspeptin mediates male hypogonadism in obesity

444and type 2 diabetes. Neuroendocrinology. 2010;91:302-307

44522) Stokes VJ, Anderson RA, George JT. How does obesity affect fertility in men - and what are the

446treatment options? Clin Endocrinol (Oxf). 2014 Aug 19. doi: 10.1111/cen.12591. [Epub ahead of print]

44723) Dhindsa S, Furlanetto R, Vora M, Ghanim H, Chaudhuri A, Dandona P. Low estradiol

448concentrations in men with subnormal testosterone concentrations and type 2 diabetes. Diabetes Care

4492011;34:1854-1859

45024) Lindsted KD, Fraser GE, Steinkohl, Beeson WL. Healthy volunteer effect in a cohort study. J Clin

451Epidemiol. 1996;49:783-790

45225) Shi Z, Araujo AB, Martin S, O'Loughlin P, Wittert GA. Longitudinal changes in testosterone over

453five years in community-dwelling men. J Clin Endocrinol Metab. 2013;98:3289-3297

67 17 68 69 70 45426) Corona G, Rastrelli G, Monami M, Saad F, Luconi M, Lucchese M, Facchiano E, Sforza A, Forti

455G, Mannucci E, Maggi M. Body weight loss reverts obesity-associated hypogonadotropic hypogonadism: a

456systematic review and meta-analysis. Eur J Endocrinol. 2013;168:829-843

45727) Mensah GA, Mokdad AH, Ford ES, Greenlund KJ, Croft JB. State of disparities in cardiovascular

458health in the United States. Circulation 2005;111:1233-41

45928) Hu B, Li W, Wang X, Liu L, Teo K, Yusuf S; INTER-HEART Investigators. Marital status,

460education, and risk of acute myocardial infarction in Mainland China: the INTER-HEART study. J

461Epidemiol 2012;22:123-9

46229) McMicheal AJ, McKee M, Shkolnikov V, Valkonen T. Mortality trends and setbacks: global

463convergence or divergence? Lancet 2004;363 :1155-9

46430) Corona G, Isidori AM, Buvat J, Aversa A, Rastrelli G, Hackett G, Rochira V, Sforza A, Lenzi A,

465Mannucci E, Maggi M. Testosterone supplementation and sexual function: a meta-analysis study. J Sex

466Med. 2014;11:1577-1592

46731) Isidori AM, Giannetta E, Greco EA, Gianfrilli D, Bonifacio V, Isidori A, Lenzi A, Fabbri A. Effects

468of testosterone on body composition, bone metabolism and serum lipid profile in middle-aged men: a meta-

469analysis. Clin Endocrinol (Oxf). 2005;63:280-293

47032) Zarrouf FA, Artz S, Griffith J, Sirbu C, Kommor M. Testosterone and depression: systematic review

471and meta-analysis. J Psy Prac 2009;15:289-305

47233) Vermeulen A, Verdonck G. Representativeness of a single point plasma testosterone level for the long

473term hormonal milieu in men. J Clin Endocrinol Metab. 1992;74:939–942

71 18 72 73 74 474Legends to figures

475Figure 1. Study flow-chart showing exclusion of subjects and distribution of study sample by gonadal status

476HG=hypogonadism; pEUG=persistent eugonadal at baseline; isHG=incident secondary hypogonadism;

477psHG=persistent secondary hypogonadism; rsHG=recovery from secondary hypogonadism

479Figure 2

480A. Predictors of incident secondary hypogonadism (sHG). Data are derived from multilevel binary regression

481models, using centre as a random-effect covariate and age (categorized into 10 years age bands), smoking

482status (current smoker – yes/no), alcohol intake (alcohol consumption for ≥5 per week vs. less), education

483level [low (compulsory education only), medium (non-compulsory education below university level) or high

484(university education)], Physical Activity Scale for the Elderly (PASE, score ≤78 and >78), Chronic

485widespread pain (yes/no), marital status (no partner, having a partner not living together or having a partner

486and living together), comorbidity (presence/absence of at least one self-reported disorder) and body mass

487index (BMI) [<25 kg/m2, 25–29.9 kg/m2, and ≥30 kg/m2] as fixed-effect predictors. Gonadal status was the

488outcome with the persistent eugonadal group being the referent category. The odd ratios are shown on a log-

489scale.

490B. Predictors of recovery from sHG. Data are derived from binary regression models, using centre as a

491random-effect covariate and age [dichotomized into age <60 and age ≥60 years (yr)], smoking status (current

492smokers – yes/no), alcohol intake (alcohol consumption for ≥5 per week vs. less), education level [low and

493medium (completed non compulsory education but lower than university level) vs. high (university

494education)], PASE (score ≤78 and >78), Chronic widespread pain (yes/no), marital status (no partner vs.

495having a partner), BMI [<30 kg/m2 and ≥30 kg/m2], and comorbidity [presence/absence of at least one self-

496reported disorder] as fixed-effect predictors. Gonadal status was the outcome with the persistent secondary

497hypogonadal group being the referent category. The odd ratios are shown on a log-scale.

498C. Predictors of incident sHG. Data are derived from multilevel binary regression models, using centre as a

499random-effect covariate and age (categorized into 10 years age bands), smoking status (current smoker –

500yes/no), alcohol intake (alcohol consumption for ≥5 per week vs. less), education level [low (compulsory

501education only), medium (non-compulsory education below university level) or high (university education)],

75 19 76 77 78 502Physical Activity Scale for the Elderly (PASE, score ≤78 and >78), Chronic widespread pain (yes/no),

503marital status (no partner, having a partner not living together or having a partner and living together),

504comorbidity (presence/absence of at least one self-reported disorder) and waist circumference (WC) [<94

505cm, 94-102 cm, and ≥102 cm] as fixed-effect predictors. Gonadal status was the outcome with the persistent

506eugonadal group being the referent category. The odd ratios are shown on a log-scale.

507D. Predictors of recovery from sHG. Data are derived from binary regression models, using centre as a

508random-effect covariate and age [dichotomized into age <60 and age ≥60 years (yr)], smoking status (current

509smoker – yes/no), alcohol intake (alcohol consumption for ≥5 per week vs. less), education level [low and

510medium (completed non compulsory education but lower than university level) vs. high (university

511education)], PASE (score ≤78 and >78), Chronic widespread pain (yes/no), marital status (no partner vs.

512having a partner), WC [<102 cm and ≥102 cm], and comorbidity [presence/absence of at least one self-

513reported disorder] as fixed-effect predictors. Gonadal status was the outcome with the persistent secondary

514hypogonadal group being the referent category. The odd ratios are shown on a log-scale.

515* p<0.05; *** p<0.0001

517Figure 3. Upper panels: Association between incidence of secondary hypogonadism and baseline

518prevalence (Panel A) or follow-up prevalence (Panel B) or incident/worsening (Panel C) of hypogonadism-

519related symptoms [adjusted for age, centre, body mass index, comorbidities, smoking habits and alcohol

520intake], using the group of persistent eugonadal subjects as referent category. The odd ratios are shown on a

521log-scale. Lower panels: Association between recovery from secondary hypogonadism and baseline (Panel

522D) or follow-up (Panel E) prevalence or recovery/improvement (Panel F) of hypogonadism-related

523symptoms [adjusted for age, centre, body mass index, comorbidities, smoking habits and alcohol intake],

524using the group of persistent secondary hypogonadal subjects as referent category. The odd ratios are shown

525on a log-scale.

526* p<0.05; ** p<0.01

79 20 80