Testosterone and Insulin Resistance
Studies in healthy men have shown an inverse correlation between testosterone and insulin levels ( Table 1 ). The Telecom study involving 1292 healthy adult men demonstrated a significant inverse relationship between levels of plasma total testosterone and insulin independent of age, alcohol consumption, cigarette smoking and plasma glucose. Even though the association was reduced to some extent by obesity, it still persisted after adjustment for body mass index (BMI) and subscapular skinfold thickness. Another large prospective population-based study of 1009 men, who were followed-up for 12 years, also demonstrated similar inverse correlations between total testosterone levels and fasting blood glucose and BMI.
In diabetic men, an early case–control study showed that androgen levels were lower than in normal men. Subsequently in a larger cross-sectional study of 985 men from the Rancho Bernardo community in California, of whom 110 were diabetic, testosterone levels were found to be lower in diabetic men, although the association was not as great after controlling for age and BMI. Of the diabetic men, 21% were hypogonadal compared to 13% of nondiabetic men, and the testosterone levels were related to the degree of glycaemia as assessed by fasting plasma glucose concentrations.
Another case–control study by the same group measured plasma androgen levels in 44 men with type 2 diabetes and compared them with 88 age-matched men with normal glucose tolerance. Men with diabetes were found to have significantly lower levels of both total and bioavailable testosterone, as well as dehydroepiandrosterone sulfate (DHEAS), than controls, even when these associations were adjusted for obesity and fat distribution. As both testicular and adrenal androgens were reduced, a selective effect of diabetes on the testicular Leydig cells was thought to be unlikely.
A larger German study comparing 155 male type 2 diabetic patients with 155 healthy controls showed that free testosterone levels were lower in type 2 diabetic men and inversely correlated with BMI. However, no correlation was found between testosterone and serum levels of C-peptide. A recent study in 103 type 2 diabetic men showed the prevalence of hypogonadism to be 33% based on assessment of free testosterone by the equilibrium dialysis method. Furthermore, men with impaired glucose tolerance and not diabetes were also found to have low total testosterone levels.
The association between hypogonadism and insulin levels in men has also been reported in studies on patients undergoing treatment for prostate carcinoma. Androgen ablation has been the mainstay treatment for metastatic disease as the growth of cancer cells in the prostate is stimulated by testosterone. Hormonal therapy consists of either surgical or medical castration induced by GnRH agonists, antiandrogens, or a combination of both. Dockery et al. showed that fasting insulin levels increased after 3 months in 16 men with prostate cancer under treatment with GnRH agonists as compared to age-matched controls. Arterial stiffness also increased in these men but there was no significant change in BMI or serum glucose. Smith et al. also found an increase in serum insulin levels and arterial stiffness, after 3 months, in 22 men in this group. These men also had an increase in fat mass and a decrease in lean mass. Similarly, in another study, 30 men who underwent surgical castration for primary prostate limited adenocarcinoma had an increase in both fasting and postprandial glucose as well as postprandial insulin, 1 month after surgery.
Ageing in males is accompanied by a progressive decline of gonadal function manifested by a fall in total, bioavailable and free plasma testosterone levels. Free and bioavailable testosterone levels decline more steeply than total testosterone levels because of the age-associated increase in SHBG level. Ageing is also associated with an increasing prevalence of type 2 diabetes. The European Group for the Study of Insulin Resistance showed that insulin action declines with ageing. Another study also demonstrated an age-related impairment in glucose handling.
The Massachusetts Male Ageing Study is a large population-based study of 1156 men aged 40–70 years who were followed up for 7 to 10 years. In this study the mean baseline total and free testosterone and SHBG levels were significantly lower among men who later developed diabetes. Similarly, another retrospective analysis from participants enrolled in the Multiple Risk Factor Intervention Trial (MRFIT) showed that the nondiabetic men who subsequently developed diabetes during 5 years' follow-up also had significantly lower levels of free testosterone and SHBG than those who did not. Further evidence that there is a prospective association between low endogenous testosterone levels and future onset of type 2 diabetes was found in older men taking part in the Rancho Bernardo study. This study reported a significant inverse relationship between low baseline total testosterone levels and follow-up levels (8 years later) of fasting and postchallenge glucose and insulin levels, as well as HOMA in men. Another study in 659 elderly men found that total and free testosterone and SHBG levels were negatively correlated with glucose and insulin values.
Further insight into the potential role of male hormones in the future development of insulin resistance and diabetes is provided by a study performed in relatives of diabetic patients. It is well established that first-degree relatives of type 2 diabetic patients have a higher risk of developing diabetes. Jansson et al. compared 33 healthy first-degree relatives of type 2 diabetic patients with 33 age-matched controls. Relatives showed decreased insulin sensitivity as measured by the euglycaemic hyperglycaemic clamp method and this difference was significant for males only. Male relatives of the diabetic patients had lower plasma total testosterone levels as compared to the male controls and the total testosterone levels observed in these relatives were positively associated with insulin sensitivity. As SHBG levels were similar between groups, it was postulated that the alterations in total testosterone levels would reflect the free hormone levels as well. Thus it would seem likely that dysregulation of androgen levels could contribute to the development of insulin resistance in male subjects who have a higher genetic predisposition for type 2 diabetes.
In summary, these studies suggest that low testosterone levels in men may potentially be a contributory factor to the development of insulin resistance and the subsequent progression to type 2 diabetes.
SHBG is the circulating steroid-binding protein produced by the liver that binds testosterone with high affinity. It is an important regulator of androgen homeostasis and functions as a modulator of androgen delivery to the tissues. SHBG concentration falls during puberty in both boys and girls. Serum SHBG levels are primarily regulated by sex steroids and thyroxine.
It has been suggested that the link between total testosterone and insulin resistance is due to the negative relationship between SHBG and insulin, with low SHBG leading to low total testosterone. Birkeland et al., using the insulinglucose clamp technique, demonstrated an inverse correlation between insulin resistance and serum SHBG levels in 23 type 2 diabetic men that was independent of serum insulin or C-peptide level as well as being independent of obesity and abdominal fat accumulation. In the Telecom study, healthy men with lower total testosterone levels had significantly higher insulin levels and markedly reduced levels of SHBG. Bioavailable testosterone levels, however, were not significantly different in the two groups, implying that the link between total testosterone and plasma insulin could be explained by the negative association between SHBG and plasma insulin. Similarly, Andersson et al. found that total testosterone and SHBG levels were significantly lower in diabetic men than nondiabetic control subjects and had a negative correlation with insulin values. There was no difference in free testosterone levels between the groups, again demonstrating that low total testosterone levels were secondary to the low SHBG. Even in relatives of hypertensive men, an inverse relationship was found between low total testosterone and SHBG concentrations and lower insulin sensitivity, with no change in free testosterone levels.
Insulin is an important regulator of SHBG production by the liver. In vitro studies have shown that insulin in physiological concentrations was a potent inhibitor of SHBG production by cultured hepatoma cells. Peiris et al. also showed a significant association between SHBG levels and the insulin secretory pulse interval but not with peripheral insulin sensitivity in 10 nondiabetic men. Furthermore, Pasquali et al. demonstrated that inhibition of insulin secretion by giving diazoxide to normal weight and obese men led to increased SHBG levels. Acute hyperinsulinaemia has also been found to result in a small but significant reduction in SHBG concentration in healthy men. Nestler has thus suggested that lower SHBG levels may be a marker for hyperinsulinaemia and insulin resistance. Men with low SHBG concentrations have an increased risk of developing the metabolic syndrome. Therefore, the available data, in men, suggest that insulin resistance maybe a determinant of SHBG levels.
Obesity is the most common cause of insulin resistance. BMI is traditionally used as an indicator of overall obesity. However, certain patterns of fat distribution are more closely related to increased incidence of diabetes and cardiovascular disease. Abdominal or central obesity, as assessed by waist/hip ratio, is an essential component of metabolic syndrome and more strongly linked to the development of impaired glucose tolerance. Visceral fat, which constitutes a significant proportion of the intra-abdominal fat, has certain characteristic metabolic and anatomical features. Visceral adipose tissue is more highly metabolically active than any other adipose tissue in the body. Furthermore, the visceral fat is drained through the portal vein to the liver, in contrast to the peripheral fat, which is drained by the systemic circulation. The above two processes result in the liver being exposed to higher concentrations of free fatty acids produced by the adipocytes than in any other organ. Free fatty acids decrease hepatic insulin binding and extraction, increase hepatic gluconeogenesis and increase hepatic insulin resistance. These effects ultimately lead to peripheral hyperinsulinaemia and systemic insulin resistance (Fig. 2).
Role of visceral fat in peripheral hyperinsulinaemia and systemic insulin resistance. Increasing abdominal fat leads to liver being exposed to higher concentrations of free fatty acids. The free fatty acids increase hepatic glucose production and decrease hepatic insulin uptake. This Results in systemic hyperinsulinaemia and skeletal muscle insulin resistance, which in turn causes further release of insulin by the islet cells.
Unlike the situation in women, in men there is an inverse relationship between serum testosterone levels and visceral fat mass. The visceral obesity in men is associated with relative hypogonadism. Obesity itself is one of the several conditions that can result in a low SHBG level. In the HERITAGE Family Study, increasing total body fat content and visceral adiposity were associated with decreased plasma levels of SHBG. As a result, total testosterone is frequently low but the free testosterone is normal, suggesting that this not a true clinical hypogonadism. This is generally seen in moderate obesity.
By contrast, other studies have shown that free testosterone levels are low in obese individuals ( Table 2 ) and the relative hypogonadism is proportional to the degree of obesity. Abdominal or upper body obesity is more strongly related to free testosterone levels than other forms of obesity. Haffner et al. found, in a population of 178 men recruited to the San Antonio Heart Study, that BMI was inversely related to total and free testosterone as well as SHBG level. Waist/hip ratio was also strongly inversely related to total and free testosterone. Similarly, Abate et al. showed that subcutaneous fat accumulation in the truncal area is highly predictive of low plasma concentrations of free testosterone. Studies by Seidell et al. and Phillips have also reported that waist/hip ratio in men was significantly inversely correlated with total testosterone, free testosterone and SHBG levels. Pasquali et al., on the other hand, found a significant inverse relationship between BMI and both total and free testosterone and SHBG but no association between waist/hip ratio and any sex hormone or binding protein.
The prevalence of obesity in ageing men has increased and is a strong predictor of the testosterone deficiency seen in ageing males. Hypogonadal men also have a reduced lean body mass and an increased fat mass. Vermeulen et al. reported in a study of 57 men between 70 and 80 years that testosterone levels correlated negatively with percentage of body fat, abdominal fat and insulin levels. Chang et al. also showed that elderly men with type 2 diabetes had higher BMI, waist/hip ratio and lower serum testosterone levels than elderly men without type 2 diabetes. Testosterone levels correlated negatively with BMI, waist/hip ratio and skinfold thickness.
The changes in total and free testosterone concentrations are reversible with weight loss. Strain et al. assessed the effect of weight loss on sex hormones in 11 healthy obese men. Weight loss of between 26 and 129 kg over 5–39 months produced significant increases in mean plasma total and free testosterone and SHBG levels. The increases in plasma free and total testosterone and SHBG levels were also proportional to the degree of weight loss. Similar Results have also been reported in other smaller studies.
The underlying mechanisms responsible for the reduced testosterone levels in obese men are unknown. The reduction in free testosterone seen in massive obesity is not accompanied by a reciprocal increase in LH, suggesting a form of hypogonadotrophic hypogonadism. One hypothesis postulated for the decreased free testosterone in massively obese individuals is functional alterations at the hypothalamicpituitary level of the testicular axis characterized by decreased amplitude of the LH pulses. Some rare hypothalamic syndromes, such as Prader–Willi syndrome, are associated with both obesity and hypogonadotrophic hypogonadism.
Another possible mechanism to explain the aetiology of low testosterone levels and the subsequent insulin resistance in obese men is hyperoestrogenemia. Earlier studies found increased serum levels of oestradiol and oestrone in obese men. This primarily occurs as a result of increased peripheral conversion of androgens to oestrogens through the action of the enzyme aromatase, which is present in higher levels in the adipose tissue as compared to other tissues. This increase in serum oestrogen concentration is, however, not accompanied by overt signs of feminization. It is thus possible that the increased oestradiol levels contribute to the insulin resistance in obese men. Phillips et al. found in 80 adult men that both total and free testosterone correlated inversely, and the ratio of oestradiol to testosterone directly, with insulin levels. However, after controlling for visceral adipose tissue, only the oestradiol to testosterone ratio and insulin concentration remained significant. Similarly, another study found higher oestradiol levels in diabetic subjects compared to the nondiabetic ones. A small study of six obese men treated with the aromatase inhibitor testolactone showed a decrease in oestradiol and an increase in testosterone levels. However, others have found no relationship between oestradiol concentrations and glucose or insulin levels or insulin resistance. Administration of ethinyl oestradiol to normal men has been reported to induce insulin resistance. There are case reports of two men, one with oestrogen resistance caused by a mutation of the oestrogen receptor alpha gene and another with low oestradiol and elevated testosterone levels as a result of a mutation in the aromatase gene. These men had glucose intolerance and insulin resistance.
At the cellular level, adipocytes express androgen receptors. Testosterone inhibits the activity of lipoprotein lipase, the main enzymatic regulator of triglyceride uptake in adipose tissue. This Results in inhibition of triglyceride uptake, increase in lipid mobilization and a subsequent decrease in visceral adipose tissue mass (Fig. 3). In the ageing male the natural diminution in testosterone is contributory to visceral adiposity. Furthermore, the relative hypogonadism produced in abdominally obese men also contributes to an increase in fat mass. Tsai et al. reported that a low baseline total testosterone level in Japanese-American men predisposed to an increase in visceral adiposity.
Interaction between adipose tissue, testosterone and insulin resistance. The enzyme aromatase, present in high concentrations in adipose tissue, converts testosterone to oestrogen. Increasing abdominal fat leads to an increased aromatase activity. The resulting low testosterone increases lipoprotein lipase activity and triglyceride uptake leading to an increased visceral adiposity and insulin resistance. This in turn causes further hypogonadism and abdominal fat deposition. Furthermore, with increasing visceral fat, LH pulse amplitude is reduced, probably through the action of certain factors such as leptin at the pituitary level, leading to further reduction in testosterone levels.
Cohen has described the hypogonadal–obesity cycle. During the hypogonadal state, there is an increase in deposition of abdominal adipose tissue. This Results in increased aromatase activity leading to a greater formation of oestradiol from testosterone. This will then lead to a further reduction in serum and tissue testosterone concentrations, increased deposition of abdominal fat and progressive hypogonadism.
Leptin is the adipocyte-secreted protein product of the ob gene. It is strongly linked to obesity and regulates weight and adipose tissue mass. Serum leptin levels correlate positively with age, BMI, serum insulin and fat mass and inversely with testosterone. Leptin levels are higher in ageing males with lower testosterone and testosterone replacement therapy corrects this. The mechanism is unclear but is likely to be related to a combination of reduction in adipose tissue mass and a direct suppressive effect on ob gene expression. As total body fat mass increases with low testosterone, hormone resistance develops for leptin and insulin. Increasing leptin fails to prevent weight gain and the hypogonadal–obesity cycle ensues, causing further visceral obesity and insulin resistance. Although the mechanisms responsible for the hypogonadism in obesity are varied, testosterone therapy in obese men reduces visceral fat mass. In elderly men, studies have demonstrated that testosterone decreases body fat mass and increases the lean mass. There is a variability in the responsiveness of the body fat to testosterone administration that depends on the duration of therapy, although other factors such as pretreatment body composition and the age of the subjects also play a role.
Clin Endocrinol. 2005;63(3):239-250. © 2005 Blackwell Publishing
Cite this: Androgens, Insulin Resistance and Vascular Disease in Men - Medscape - Sep 01, 2005.