The topic that I will discuss today is the interplay between obesity and androgens in the pathogenesis of the metabolic syndrome.
Metabolic syndrome is a cluster of risk factors and obesity is a major component. Obesity is also a risk factor for androgen reduction, and androgen reduction itself can cause an increase of fat mass and obesity in males. It is also possible that androgen reduction is associated with the development of the metabolic syndrome.
In the first part I will try to assess how male obesity is involved in the reduction of androgen production. The reduction is surely mediated by mechanisms that occur at the level of the hypothalamic-pituitary axis through the increase of circulating oestrodial, due to the aromatisation of testosterone at augmented deposits of fat tissue, and consequent reduction of LH pulses. But fat is also a tissue that produces leptin, which acts as an anti-testosterone agent; with increasing adiposity there is a reduction in insulin sensitivity which leads to higher insulin and lower levels of sex hormone binding globulin (SHBG). All these conditions can lead to androgen reduction.
In the late ‘90s we demonstrated that leptin is a major contributor of androgen reduction in obese subjects. In that study we examined 38 men of different levels of obesity and we found that LH-induced testosterone production was blunted in obese subjects and that the reduction was inversely related to the degree of adiposity. In addition, the amount of total testosterone of the subjects was inversely related to circulating leptin. This study was cross-sectional and not mechanicistic, but clearly suggests that leptin is involved in the pathogenesis of low androgens of obese subjects. At the level of the Leydig cell this is further proved by the fact that Leydig cells have functional leptin receptors; when Leydig cells are incubated with leptin and stimulated by hCG, there is a leptin-induced dose-dependent inhibition of testosterone production. The IC-50 of leptin inhibitory effect on hCG-stimulated testosterone production is around 10 ng/ml leptin, which is consistent with the concentration usually present in obese subjects.
In a very recent paper an American group showed that low androgenicity in obese subjects is mainly due to reduced insulin sensitivity. In their model they eliminated the confounding effect of the hypothalamus through the use of a GnRH antagonist, concomitantly injecting the subjects with GnRH or hCG. They showed that after GnRH stimulation there was no difference in LH production between insulin resistant and non-insulin resistant obese subjects, while the hCG-induced testosterone production was blunted in patients with insulin resistance. The results of this study appear to indicate that insulin resistance affects mostly the Leydig cell production of testosterone. This is also showed by the following slide in which the authors found in the same study that the glucose disposal rate, which is an index of insulin sensitivity, was directly related to the amount of hCG-induced testosterone production.
According to these findings, we re-analysed the data of our previous studies by multiple regression analysis and found that fat mass explained most of the testosterone reduction in insulin resistant subjects. The additional contribution of HOMA index, which as already explained by Dr. Jones is an index of insulin sensitivity, on testosterone reduction was only 2%, indicating that fat mass and not insulin resistance is the major contributor of androgen reduction in obese subjects.
But androgen reduction is also involved in changes of body composition as we already heard.
There are several studies which indicate that testosterone is a strong modifier of body composition. We just published a meta-analysis, in which we analysed a total of 29 randomised controlled trials (RCT) evaluating the effects of testosterone administration to middle-aged and ageing men on body composition, muscle strength, bone density, markers of bone metabolism and serum lipid profile. Overall, 1083 subjects were evaluated, 625 randomized to testosterone, 427 to placebo and 31 to observation (control group). Weighted mean age was 64·5 years (range 49·9–77·6). Testosterone treatment produced: (i) a reduction of 1·6 kg of total body fat, corresponding to 6·2% variation of initial body fat, (ii) an increase in fat free mass of 1·6 kg, corresponding to +2·7% increase over baseline and (iii) no change in body weight. These effects are in the range of what has been already shown by the first speaker and are similar to the observations of an another not RCT by Swerdloff et al. In the latter study, testosterone gel was given for six months and as you can see lean body mass increased and fat body mass clearly decreased, with minimal changes in total body mass.
One important point is derived from experiments performed in a human model of Leydig cell clamp, in which testosterone production by Leydig cells is abolished through a GnRH agonist and subjects are replaced with testosterone esters at increasing concentrations. In this way different plasma testosterone concentrations are achieved in plasma and it is possible to identify which is the testosterone level at which there is an effect on body composition. According to this study, you can see that both the 25 mg and 50 mg of testosterone doses injected i.m. weekly for 16 weeks lead to testosterone levels steadily below 320 ng/dl and to a marked increase in total adipose tissue; the increases of fat mass were similar between appendicular and trunk body-sites. This study was performed in healthy young individuals, and demonstrates that in these subjects when circulating testosterone is below 320 ng/dl there is a clear cut change in body composition, with more fat both in trunk and the appendices. However, these were young subjects. In older men, it has been shown that long-term androgen administration causes a more prevalent reduction of visceral fat compared to subcutaneous fat, indicating that in aging men with low androgens, testosterone replacement therapy mostly affects visceral fat.
The reduction in androgen can cause impaired differentiation of the mesenchimal stem cell.
This derives from nice, basic science studies. I wish to show a very recent one, in which the authors incubated 3T3-L1 pre-adipocytes with testosterone and followed the cell differentiation into mature adipocytes. Red oil is a marker of adipocyte differentiation and you can see that testosterone markedly inhibited cell differentiation. Androgen receptor is expressed in these pre-adipocytes and the incubation of the cells with testosterone or DHT, caused a translocation of the receptor from the cytoplasm to the nucleus, with a strong inhibition of two transcription factors, which are very important for the commitment of pre-adipocytes into mature adipocytes, i.e. C/EBP- a and PPAR- g .
Androgens have the ability to inhibit the commitment of the precursor cell into the adipogenic lineage while favour maturation into myocytes.
In another study, which was two years prior to the last one, the same authors found that when a pluripotent cell is incubated with testosterone there is a marked stimulation of a marker protein of the myogenic lineage, i.e. cells develop abundant levels of myosin heavy chain. Also in these cells, testosterone inhibits PPAR- g , which is the transcription factor that directs the cells into the adipocyte lineage.
One possible effect of testosterone that we thought of, was that testosterone can affect an enzyme that is 11- b -HSD-type 1, which is known to convert cortisone to cortisol. The hypothesis was that testosterone could inhibit the expression/function of this enzyme leading to low cortisol in abdominal fat thus reducing the susceptibility to abdominal obesity. However at present this hypothesis is not supported by experimental findings; in fact, as you can observe in the slide, we found that while insulin stimulated cortisone to cortisol conversion in 3T3-L1 adipocytes, androgens had no measurable effects.
What about low androgen and adipokines? Dr. Jones piloted a study on this topic. We know that obesity is associated with macrophage accumulation in adipose tissue. This kind of accumulation causes a state of inflammation. These macrophages have the ability to produce cytokines and TNF-a is one the major inflammatory cytokines. TNF-a is also an anti-insulin agent. In the study already shown by Dr. Jones from his group, you can observe that one month of testosterone therapy leads to a marked decrease in circulating TNF-a , with a concomitant increase in IL-10 which is a anti-inflammatory cytokine. The results of this study demonstrate that testosterone treatment has a favourable effect on the cytokine profile.
What about the metabolic syndrome then?
There are observational studies, in which it is addressed the issue of the contribution of age and the decline of androgen levels to the features of metabolic syndrome in men. This is just one of the studies showing that in aged subjects in the lower tertiles of testosterone there is an increase in the number of metabolic syndrome features. In contrast when testosterone is high, there is a 5-folds reduction in the occurrence of metabolic syndrome (from 44.2 to 8.9%).
This is a nice study from Maggi et al. performed in more than 1,000 subjects, showing that in diabetic patients compared to non-diabetic patients there is a higher number of low testosterone compared to non-diabetic subjects; the group that was mostly affected was the group between 50-59 age years. One marker of androgenicity that we must think of is PSA, and in the subjects of this decade with low testosterone there was a marked decrease of PSA. So, when we observe a patient with low testosterone but not adequately low PSA, we must pay particular attention in the diagnostic work up and follow him carefully during eventual androgen replacement.
Another important study was performed on the association between levels of endogenous sex hormone and the occurrence of the metabolic syndrome in aging men, The authors showed that for one standard deviation increase in total testosterone, there was a strong reduction in the risk to develop metabolic syndrome. This phenomenon was true both in crude population, or when numbers were adjusted for age, smoking, alcohol consumption, physical activity, BMI, and waist circumference. So, the concept that emerges from this study is that only one standard deviation increase of testosterone is needed to decrease the risk of metabolic syndrome.
A very recent prospective study by McKinlay and his group, which was just published this year (2006) in the January issue of JCEM, found that in the Boston population area, looking at subjects without metabolic syndrome at baseline, between 1987 and 1989, which were followed up to 2004, the risk to develop metabolic syndrome was much higher when the subjects had low testosterone levels. More important, when subjects with BMI less than 25 were compared with subjects with BMI more than 25, the risk to develop metabolic syndrome was higher in subjects who were not obese (relative risks of 2.5 compared to 1.22).
The fact that the Authors did not find an increased risk of metabolic syndrome with low SHBG or total T in men with a BMI≥25 suggests that in overweight and obese men, adiposity is the dominant risk factor for developing metabolic syndrome and supports a major role for exercise and weight loss to prevent the development of metabolic syndrome.
What about interventional studies?
There are not a lot of interventional studies, as it has been stated by the previous speakers. A very initial one was performed in 1993, in which testosterone gel was administered for nine months in subjects with low normal testosterone. There was an increase in glucose disposal rate, a reduction of fasting blood glucose, serum cholesterol and serum triglycerides.
This is a small study, but it is interesting in the sense that 18 selected men with low testosterone were treated with an androgen gel aimed to obtain a through level of testosterone between 4 and 10 ng/ml. In this study testosterone treatment was capable to reduce fasting plasma insulin, fasting plasma glucose, the HOMA index and leptin levels.
One important consideration is that when we think of replacing androgen in subjects with low androgens, it is likely that the threshold of androgen action on target tissues is different. So in this slide I tried to depict what can be the case. Below 300 ng/dl, testosterone action on libido appears to be preserved, and just a little bit above 300 is needed to get a good effect on erectile function and possibly bone density.
But when the clinical target is a change in body composition, the level of plasma testosterone needed is that which is normally present in a young population, i.e. around 500 ng/dl. This is the level which is not the one that is really recommended right now for replacement therapy because the recommendation is to be just above 300. But if we need to have a complete effect of androgen replacement therapy, we must achieve the range of young adults. Of course, the risk of testosterone side effects is increased at higher circulating testosterone levels.
So, we can conclude that fat mass and insulin resistance are both associated with primary hypogonadism; however, from ours and other authors’ data it seems that fat mass, leptin and possibly other adipokines exert the predominant effect. In turn, low testosterone itself may favour insulin resistance, by increasing fat mass and reducing muscle mass; to exert this effect the testosterone cut off value appears to be below 320 ng/dl,. according to the Leydig cell clamp study of Bhasin et al. Low testosterone would determine also a shift of cytokine profile toward an enhanced production of pro-inflammatory peptides, like TNF-a .
Observational and interventional studies indicate that testosterone therapy improves insulin resistance. Testosterone might reduce the risk of metabolic syndrome, this by acting on body composition, lipid profile, the shift of pro-inflammatory cytokines and possibly other effects. In order to prove this important aspect of testosterone replacement therapy more and more studies are needed, and I am confident that they are ready to be performed.
Thank you for your attention.
Copyright © 2006 ISSAM