Such beneficia therapeutic effects on CV risk factors now need to be confirmed in larger, longer term trials, as does safety. The diagnosis of clinically relevant hypogonadism will be discussed in the light of the recent evidence and current clinical guidelinesLow circulating levels of testosterone are frequently found in men with the metabolic syndrome and T2DM.
Such beneficia therapeutic effects on CV risk factors now need to be confirmed in larger, longer term trials, as does safety. The diagnosis of clinically relevant hypogonadism will be discussed in the light of the recent evidence and current clinical guidelinesLow circulating levels of testosterone are frequently found in men with the metabolic syndrome and T2DM. There is increasing recognition that this hypotestosteronaemia is associated with important clinical associations with CV risk factors such as visceral obesity, insulin sensitivity, dyslipidaemia, inflammatory state and intima media thickness (IMT), a surrogate marker of atheroma. A prospective study has found that low testosterone levels were associated with increased mortality over a follow-up period of eight years.1 Whether or not this testosterone deficiency is a risk factor for, or a consequence of, metabolic syndrome and subsequently diabetes is not fully clear, but research suggests it is a combination of both factors. The important question is whether or not TRT will ultimately have a positive effect on these CV risk factors by first reducing progression, and second, leading to some resolution of the atherosclerosis.
This review will discuss the current evidence for a role of testosterone in the pathogenesis of these conditions and examine early evidence on the potential benefits of treating these men with TRT. To be able to understand the role of testosterone in this process, and to decide if an individual requires testosterone substitution, it is important to first review the physiology and clinical diagnosis of hypogonadism
In the circulatory system testosterone exists in three major fractions – free (fT), bound to sex hormone binding globulin (SHBG) and albumin – which together provide the total testosterone. Testosterone is tightly bound to SHBG, dissociating slowly and is considered to be biologically inactive, whereas it is weakly bound to albumin and dissociates readily. The sum of the fT and albumin-bound testosterone is known as the biologically active or bioavailable testosterone (bT). Evidence exists that shows fT and the bT have stronger correlations with markers of androgenisation than the total testosterone. In routine clinical practice the total testosterone is usually measured. It is important to recognise that fT measured using any of the analogue assays are partly SHBG-dependentSerum levels of SHBG are affected by several factors. For example, levels are found to be lower in association with obesity, insulin resistance, atorvastatin therapy and tobacco smoking, and higher with ageing These effects need to be taken into consideration, particularly in patients with metabolic syndrome, diabetes and cardiovascular disease (CVD).Testosterone has a diurnal rhythm with levels peaking between 06:00 and 08:00h and reaching a nadir between 18:00 and 20:00h. Blood for testosterone assays should be taken before 11:00hDiagnosis of Hypogonadism
The most recognised terminology for a symptomatic testosterone-deficient state by medical practitioners is hypogonadism. The definition of hypogonadism is a clinical syndrome that comprises both symptoms, as well as biochemical evidence of testosterone deficiency. However, there is no consensus as to the level below which an individual is considered to be hypogonadal.2 There has been a plethora of guidelines produced for the diagnosis and management of hypogonadism. Of these, there are two major guidelines. The Endocrine Society suggests that <10.4nmol/l is consistent with the diagnosis; however, not all pane members agreed.3 The International Society of Andrology (ISA), the nternational Society for the Study of the Aging Male (ISSAM) and the European Association of Urology (EAU) guidelines for late-onset hypogonadism (LOH) state that total testosterone <8nmol/l requires testosterone substitution, whereas >12nmol/ does not.4 In the presence of symptoms levels between 8 and 12nmol/l, patients should be considered for a therapeutic trial of testosteroneIn men with metabolic syndrome, diabetes or CVD, total testosterone may not be the ideal test for assessment of testosterone deficiency due to common factors to these conditions that affect the SHBG level. Tota testosterone is a good screening test but is not as sensitive at detecting biochemical testosterone deficiency as fT or bT, following research by Malkin et al., in men with CVD.5 Mathematical calculations for fT and bT are available, but should be validated in individual laboratories.6
Simvastatin therapy reduced CVD risk by 22% in the Heart Protection Study (HPS),7 and atorvastatin therapy reduced CVD risk by 32% in the Collaborative Atorvastatin Diabetes Study (CARDS).8 Nevertheless, significant residual cardiovascular risk remained in diabetic patients treated with statins in HPS and CARDS (78% and 68%, respectively).
Residual CVD risk remains in all patients treated with statins; however, residual CVD risk is particularly high in patients with diabetes treated with statins. Even after patients with diabetes were treated with statins, their CVD event rates (i.e. residual CVD risk) in large-scale clinical trials were higher than the CVD event rates of those patients without diabetes on placebo.Thus, statins reduce but do not eliminate the increased CVD risk associated with diabetes. HPS = Heart Protection Study (simvastatin),7 CARE = Cholesterol and Recurrent Events (pravastatin),24 LIPID = Long-Term Intervention with Pravastatin in Ischaemic Disease (pravastatin),25 PROSPER = Prospective Study of Pravastatin in the Elderly at Risk (pravastatin),6 ASCOT-LLA = Anglo-Scandinavian Cardiac Outcomes Trial-Lipid Lowering Arm (atorvastatin),26 TNT = Treating to New Targets (atorvastatin 10 mg versus 80 mg).27
*CHD death, non fatal MI, stroke, revascularizations
†CHD death, non fatal MI, CABG, PTCA
‡CHD death and non fatal MI
§CHD death, non fatal MI, stroke
CHD death, non fatal MI, resuscitated cardiac arrest, stroke
Several studies have found that men with diabetes have a higher prevalence of low testosterone levels compared with matched, healthy controls.7–11 The Rancho-Bernado study, the largest of its kind, found that bT, as well as total testosterone, were lower in the diabetic group.8 mportantly, this has demonstrated that bT and fT (measured by equilibrium dialysis) levels were consistent with testosterone deficiency in one-third of the diabetic men.8,12<,13 These methods exclude any confounding effect of SHBG levels due to insulin resistance. Two studies have shown that testosterone levels are also low in men with metabolic syndrome.14,15Kapoor and colleagues assessed the prevalence of clinical hypogonadism in 355 men with T2DM.13 Previous studies had reported an association only with biochemical levels of testosterone and not the presence of symptoms, which are an essential component of the diagnosis. In this study 17% had total testosterone of <8nmol/l with symptoms, and 14% had a bT <2.5nmol/l (less than the lower limit of the normal range). Afurther 25% had total testosterone between 8 and 12nmol/l. Importantly, 26% had primary hypogonadism and 10% hypogonadotrophic hypogonadism (luteinising hormone (LH) <2U/l). The remainder had normal LH levels consistent with the findings of Dhindsa et al.12
Low Testosterone – A Risk Factor for Metabolic Syndrome and Diabetes
The Multiple Risk Factor Intervention Trial (MRFIT), the Massachusetts Male Aging Study (MMAS) and a study from Finland have all reported that lower levels of total testosterone, and fT in the case of the MMAS, as well as SHBG, are independent risk factors for the subsequent onset of diabetes. 15–17 The MMAS also has reported that these parameters are ndependent risk factors for the development of metabolic syndrome Additionally, the MMAS has reported that in non-obese men low testosterone is a risk factor, providing evidence that androgen deficiency predisposes to the pathogenesis and is not only a consequence of obesity.19 This finding is strongly supported by a recent report of a population of 1,413 men from the Third National Health and Nutrition Survey (NHANES III).20 This demonstrated that men in the lowest tertile of either fT or bT, but not total testosterone, were approximately four times more likely to develop diabetes compared with those in the third tertile after adjustment for adiposity, age, race and ethnicityTestosterone and Insulin Resistance
Insulin resistance is a central abnormality in the pathogenesis of the metabolic syndrome and T2DM. Visceral obesity, lack of exercise and genetic defects all contribute to the development of insulin resistance nsulin resistance in turn is associated with glucose intolerance, dyslipidaemia, hypertension, endothelial dysfunction and microalbuminuria, a hypercoaguable state and elevation of pro-inflammatory proteins such as C-reactive protein (CRP) and cytokines. All of these factors are involved in the pathogenesis of atherosclerosis. By extrapolation, any change in lifestyle or therapy that reduces insulin resistance should lead to a reduction in CVD risk. Several studies have demonstrated an association between testosterone and insulin sensitivity.21The Telecom study found an inverse correlation between testosterone and insulin levels in 1,292 healthy adult men that was independent of age, cigarette smoking, alcohol intake and blood glucose.22 This association, although slightly less positive, also persisted after adjustment for body mass index (BMI). TRT in hypogonadal men and in men with obesity improves insulin sensitivity.23 Beneficial effects of testosterone treatment on insulin sensitivity have also been demonstrated in men with HIV infection and in chronic heart failure patients, both states associated with insulin resistance.24,25
There has been only one study where TRT has been given to hypogonada men with T2DM to investigate the effect on insulin resistance.26 This was a double-blind placebo crossover study of seven months duration, which ncluded a one-month washout between placebo and treatment phases The treatment involved 200mg Sustanon® testosterone esters administered intramuscularly every two weeks. Testosterone treatment resulted in significant improvement in insulin resistance as assessed by homeostasis model assessment-insulin resistance (HOMA-IR) (mean -1.73), fasting blood glucose (-1.58mmol/l), HbA1c (-0.37%, from 7.28 to 6.91%) and total cholesterol (-0.4mmol/l). There was also a reduction in waist circumference, but no change in percentage body fat or BMI Furthermore, 10 insulin-treated men in addition to a lowering of HbA1c had to have their insulin dose reduced as a result of increased frequency of hypoglycaemia. One previous study of 48 poorly controlled diabetic men which was non-blinded, reported mean fall in HbA1c of 1.8% (from 10.4 to 8.6%).27 A small study of only 10 men did not demonstrate any effect of testosterone on glycaemic control.28These pilot studies, albeit relatively short-term, do suggest that testosterone substitution does have beneficial effects on glycaemic control, and by extrapolation should decrease risk of both micro- and macrovascular diabetic complications based on the UK Prospective Diabetes Study (UKPDS) findings. However, these potential benefits can be answered only by larger and longer term studies. One such study is currently underway. TIMES2 (Testosterone replacement In MEtabolic Syndrome and type 2 diabetes) is a large, multicentre, European study over two years using a novel metered-dose 2% testosterone gel. The primary outcome is the effect of TRT on insulin resistance with secondary outcomes that include glycaemic control, lipid profile, waist circumference and blood pressure at six months and at two years. Two-year data will be analysed for a trend in reduction of overall CVD risk.Cardiovascular Risk Factors
Visceral obesity is a major independent CVD risk factor, as clearly demonstrated by the INTERHEART study.29 It has been known for severa years that testosterone plays a major role in controlling body composition, promoting increased lean mass and reduced fat mass Reduction in fat-free mass and increases in fat mass are well known to be associated with the hypogonadal state. Testosterone substitution results in normalisation of body composition, with the effects taking up to two years to maximise.30 Visceral obesity is an essential part of the metabolic syndrome. The study described above in men with T2DM showed that testosterone reduced waist circumference and also decreased levels of the adipocytokines leptin and adiponectin.1331 Similar findings occured in hypogonadal men replaced with testosteroneIt is not fully understood whether or not obesity is a cause or consequence of testosterone deficiency. It is likely to be a combination of both aetiologies. On the one hand, it is understood that as described above hypogonadism per se causes increases in the fat mass. However, adipose tissue metabolises testosterone to oestradiol by the action of aromatase, which promotes greater fat deposition, (see Figure 3).
This hypothesis was put forward as the hypogonadal-obesity cycle.31 This is further exacerbated by the inhibitory effects of oestradiol, leptin resistance and nflammatory adipocytokines on the hypothalamic-pituitary gonadal axis mpairing LH secretion. This would account for normal or low levels of LH being detected in obese patientsApart from insulin resistance and glycaemia, testosterone deficiency is also associated with dyslipidaemia, hypertension and pro-thrombotic and pro-inflammatory states; all of these conditions are known to be a consequence of insulin resistance.33 Several studies supported by a meta-analysis show that hypogonadism is associated with higher total cholesterol, low-density lipoprotein cholestrol (LDL-C), triglycerides and ipoprotein a, and low high-density lipoprotein cholestrol (HDL-C) Studies on the effect of TRT are conflicting; however, a meta-analysis did show that testosterone reduced total and LDL-C but causes a small but significant lowering of HDL-C.34 In hypogonadal men with diabetes, testosterone lowered total cholesterol, but had no effects on other components of the lipid profile over three months.35 In healthy and diabetic men, HDL-C is positively associated with testosterone.36,37 The initial fall in HDL-C may be caused by increased liver metabolism. Longer term studies are needed to establish whether or not levels rise in time.
Fenofibrate significantly reduced the incidence of non-fatal myocardial infarctions (MI), total CVD events (secondary end-point (EP)), and coronary revascularizations.The incidence of CHD events (primary EP) was reduced by 11% (p = 0.16).14
In FIELD, 78% of the study population had no prior history of CVD (n = 7,664). In this primary prevention population, fenofibrate significantly reduced the incidence of the primary end-point (CHD events) by 25% (p =
.014) and significantly reduced the incidence of the secondary end-point (total CVD events) by 19% (p = 0.004).14
Hypertension treated or untreated is associated with lower testosterone evels; however, the majority of studies have not demonstrated any beneficial effect on blood pressure. One study did find a small fall in diastolic pressure. Low testosterone is associated with a hypercoaguable state with increased plasminogen activator type 1 (PAI-1), fibrinogen and factor VII and low levels of tissue plasminogen activator (tPA).33 No studies have shown a beneficial effect of testosterone replacementTRT in hypogonadal men results in a suppression of the pro-inflammatory cytokines – TNF-cx (tumour necrosis factor-alpha) and IL-1a (interleukin-1 alpha 1) – and increased levels of the antiinflammatory and antiatherogenic cytokine IL-10.35 In men with T2DM, low testosterone evels are negatively correlated with IL-6 and CRP; however, testosterone substitution had no effect on these substances or TNF-cx.31 However, Corrales and co-workers have shown that testosterone suppresses the ex vivo production of IL-1(3, IL-6 and TNF-cx in antigen-presenting cells.38 These findings support an antiinflammatory and a potentia antiatherogenic action of testosteroneErectile dysfunction, which has a very high prevalence in diabetic men, is not only associated with CV disease, but also with testosterone deficiency in this group.39 Testosterone levels are inversely correlated with the severity of erectile dysfunction.39 Testosterone substitution can lead to an improved response to phosphodiesterase type 5 inhibitor (PDE5) inhibitors.40
*Progression of albuminuria was defined as the proportion of patients who progressed either from normoalbuminuria to microalbuminuria or from microalbuminuria to macroalbuminuria. Fenofibrate significantly improved at least two indicators of microvascular disease. Patients treated with fenofibrate (F) experienced a significantly lower incidence of laser treatment for retinopathy (p = 0.0003) and a significantly reduced progression and increased regression of albuminuria (p = 0.002), compared with those patients treated with placebo (P).14
Testosterone and Atherosclerosis
MT is a surrogate marker of atherosclerosis and is readily measurable by Doppler ultrasound in the carotid artery. Three studies have demonstrated an inverse relationship between IMT and testosterone levels.41–44 These studies have been performed in three different populations: men with diabetes, men with atherosclerosis and when ndividuals with CV disease were excluded. In a four-year follow-up study, men with the lowest testosterone at baseline showed more marked progression of the degree of IMT.45 In a study of elderly men, the degree of aortic atherosclerosis assessed by calcification on a plain X-ray negatively correlated with serum testosterone.46In over 35 cross-sectional studies, coronary heart disease (CHD) ranged from myocardial infarction (MI) through to CHD.33 Testosterone has mainly been assessed by measurement of total testosterone; however, a few studies assayed for free and bioavailable testosterone. Half of the studies demonstrated a correlation of low testosterone with CHD, whereas the rest found no association. The majority of the studies that did show the correlation measured fT of bT. Only one study has compared testosterone levels in men with significant coronary disease (>75% stenosis of at least one artery) with a control group of men with no disease at coronary angiography.47This showed a correlation between bT and CHD, but not total testosterone. Longitudinal studies, however, have not reported any influence of baseline testosterone with CHD.33In animal studies, castration is associated with the development of early vascular wall changes of atherosclerosis.48 Testosterone substitution not only retards progression of the disease, but can also ameliorate plaque formation.48 The testicular-feminised mouse (tfm), which expresses a non-functioning androgen receptor, develops aortic lipid streak formation on a high-cholesterol diet, whereas its ittermate control does not.49 However, the castrated wild-type mouse develops the same degree of lipid deposition as the tfm. If the tfm, which has low circulating testosterone levels, is given physiologica testosterone replacement, it protects against lipid deposition. This beneficial effect is mediated in part by a testosterone effect independent of the classic androgen receptor, and in part through testosterone conversion to oestrogen
There is now substantial published evidence that testosterone deficiency in men is associated with metabolic syndrome, T2DM and CVD. A high proportion of these patients have clinically relevant hypogonadism and TRT should be considered. Furthermore, longitudinal population studies have demonstrated that low testosterone levels are an ndependent risk factor for the subsequent development of metabolic syndrome and T2DM, independent of obesity. Low circulating testosterone levels positively correlate with several well-known CV risk factors. Early short-term proof-of-concept pilot studies have found that TRT improves insulin resistance, glycaemic control, waist circumference, hypercholesterolaemia and cytokine profile. Larger, longer term studies currently underway will determine if these beneficial effects persist and thus reduce CV risk.