Right, I am going to talk about diabetes, which is the far end of the spectrum of Metabolic Syndrome. And I think if we are going to be using testosterone anywhere as replacement therapy we will be using it at that end rather earlier than when we discuss treatment in Metabolic Syndrome.
Now, diabetes is a major problem in the western world and as you may know 75% of people with type II diabetes, men and women, the ultimate event that will cause them to die, in 75%, is cardiovascular disease.
Visceral obesity is very common in type II diabetes and is a major independent risk factor for coronary heart disease. And as you know, it is very much associated with insulin resistance.
What is insulin resistance? It is simply a reduced ability to respond to insulin but it is the core defect in most patients with type II diabetes, and it is, as you know, a strong predictor for the eventual development of type II diabetes in non-diabetic individuals. And as I said, it is an independent predictor of coronary heart disease, even in patients without diabetes.
Interestingly, when you look at the data, smoking, which is the characteristic high risk factor for cardiovascular disease, has the same odd ratio for coronary heart disease, as does insulin resistance. Now we are tackling this by reducing smoking but the only way we can tackle insulin resistance currently is by weight reduction, exercise and using insulin sensitisers, such as metformin and the glutazones.
So, insulin resistance sits here. The causes of insulin resistance are genetic, obesity, primarily visceral obesity, and lack of exercise. But through this pathway of insulin resistance it leads to all of these problems: hyperglycaemia, hypertension, dyslipidaemia, increased thrombotic risk, and microalbuminurian diabetics, which is closely linked to endothelial dysfunction and the mortality for someone with microalbuminuria at four years, is 33%, so that is bad news. All these come together to give this increased cardiovascular risk.
If you look at these in more detail, abdominal obesity, there are all these other factors, which have been shown to be associated with insulin resistance. With abdominal obesity I just want to draw your attention to tumour necrosis factor alpha (TNFa), and I will talk briefly about that later on. Inflammation is a key factor, endothelial dysfunction.
The major tissues, as you know, involved in insulin resistance are the liver, the muscle, and the adipose tissue. But the adipose tissue is increasingly being recognized as probably the most important organ in insulin resistance.
You are aware of all the studies that show that healthy men with low testosterone levels predicts later-onset of diabetes, such as the Massachusetts Male Aging study, MRFIT, Rancho-Bernardo and Tibblin study, which all showed that these free testosterone, total testosterone levels, if they are low they will be more likely to develop diabetes in the future. So which is the chicken, and which is the egg in this situation? Thus the evidence is suggesting that testosterone may be a contributor to the development of diabetes, if it is low.
This is the data from the Massachusetts Male Aging study, where free testosterone is as strong a predictor for the subsequent development of diabetes as is hypertension, heart disease and BMI.
So, what about diabetics? What studies are being done looking at type II diabetics?
These are a handful of studies, which have looked at diabetic compared with healthy populations, and these go back to 1984 with total testosterone being measured at that point. In 1992 bio available testosterone was measured, and then free total testosterone.
In 2004 in JCEM, Dhindsa et al actually used the equilibrium-free dialysis method of measuring free-T, which is the gold standard. They found out that 33% of men with type II diabetes were at hypogonadal levels.
What we have done in Barnsley, which is a town just north of Sheffield, with 225,000 individuals with a very high prevalence of diabetes, in some areas in Barnsley it is up to 6%, and the population of Barnsley is 98% Caucasian, so it is a unique population. We looked at 355 men who were recruited from the Diabetic Retinopathy Screening Clinics, so they were a good sample of patients from the general population, and we recorded all their demographic data and we measured total testosterone, actual bio available testosterone using ammonium sulphate precipitation, and we used our calculation of Bio-T, which was based on looking at over a 1,000 men with heart disease, the Morris formula, and also the Vermeulen formula, and just had a look to see how many patients were actually hypogonadal with low testosterone levels.
This gives you an idea of the treatments the patients were on, 70 were on diet alone, 190 were on drugs, 95 were on insulin or insulin-metformin. So that was a reasonable cross-section of a diabetic population.
The demographic data, sorry this is a complicated slide, but I just wanted to point out, that the HbA1c in this group is 7.22%. So, they were a reasonably well-controlled group of individuals. The mean total testosterone was 12.72, so that was lower than the general population. And the mean bio-available testosterone was 4.03. Our calculation here got a mean of 4.01, so verifying the accuracy of our calculation. The Vermeulen formula had a mean of 0.274, nmol/L. The blood pressure in this group was the only other thing that was quite well controlled with a mean of 143/82.
So this was the raw data where looking at total testosterone and also the free testosterone. And so the patients here in the green are total testosterones of less than eight and these are the decades. Now you’ve got a big blip here between 30 and 39, where there is almost a 30% incidence of total testosterones of less than eight. I’ll come back to that in a minute. You can see that total testosterone is very high with less than 12. And then it comes right down in this decade and then gradually increases in all parameters. Now, this particular group were very obese. These were young, very obese men with BMI’s of 45 and above. Now, would you put all of these men with a total testosterone of less than 8 on testosterone? That is the question.
When you look at the symptoms, this is the total testosterone again, just to concentrate on total testosterone, in that group of people, only about 8% have symptoms. So I think it is very important that the diagnosis of hypogonadism in studies is based on both bio-chemical and clinical analysis. You will see that the total testosterone increases as the years go on, higher incidents of less than 8, and also less than 12. So when you are above 70, 52% have a total testosterone of less than 12, and were symptomatic of hypogonadism.
If you look at the bio-active forms of testosterone, again you see this rise with age; this is the bio-available testosterone, less than 2.5, which is significantly hypogonadal. The bio-available of less than 4, which is, some patients are hypogonadal with symptoms, and some are not, and the Vermeulen formula, which mirrors quite closely from decade 40 upwards, the bio-available of 4 nmol/L in incidence. But you can see when you get past 70, when diabetic; you are very likely in a man to have symptoms with a low testosterone.
Putting everything together with the prevalence data we found that of our population of 355 men, 17.18% had a total testosterone of less than 8; 42.25% less than 12, and bio-availables less than 2.5%, 40.37 and; less than 4, 43.6%. This again matched the figure 41.9% with the Vermeulen so all of these are matching quite remarkably well. Total testosterone is as good as these markers of bio-activity and actually picking patients up with symptoms, however, when you go into the actual analysis these measures are better at picking up the borderline hypogonadal patients.
Interestingly, when you look at the causes of hypogonadism in our population we found that 26% had primary hypogonadism, with several patients with missed Klinefelter’s Disease, 10% had hypothalamic pituitary dysfunction. The majority had this mixed hypogonadism with normal LH and FSH.
We have looked at a lot of data but I just want to bring out the correlation between the bio-available testosterone and waist circumference which came down to a value of p<0.001, so essentially, the greater the visceral obesity, the lower your testosterone. It was positive for total testosterone as well. But it was much more positive for bio-available.
The point is the significant number of men with type II diabetes have symptomatic hypogonadism and 36% have classical hypogonadism and there are a lot of clinics that are not looking for these patients routinely. Testosterone levels correlate very strongly with visceral obesity in the diabetic population.
So, the next question, should we be treating these patients with testosterone?
There have been very few interventional studies of testosterone replacement therapy in type II diabetes. There was one that was published in 2003 by Boyanov et al; it was a non-blinded study of 48 men, half of whom were given oral testosterone and half placebo. They showed a significant reduction in HbAlc from a mean of 10.4 down to 8.6%. So this is a remarkably poorly controlled population getting benefit. They showed an improvement in BMI. A small study by Corrales et al had no effect of testosterone replacement therapy, this was only ten men. There have been no studies until now on insulin resistance in diabetes.
Insulin resistance is something that is not easily measured in the clinic. The gold standard is the euglycaemic clamp technique, which measures insulin resistance under hyperinsulinaemic conditions. However, the HOMA, which is the homeostasic model of insulin sensitivity, or insulin resistance, estimates insulin sensitivity in the fasted state and it has been shown to correlate very well with the euglycaemic clamp. Essentially you take three blood samples, five minutes apart, and you need the fasting insulin and fasting glucose, multiply them together divide by 22.5.
We did a double-blind placebo cross-over study to investigate the effect of testosterone replacement therapy on insulin resistance and glycaemic control in hypogonadal men with type II diabetes. So they are all symptomatic. Of the 24 men, 3 were on diet, 11 had oral treatment and 10 were on insulin. We couldn’t, obviously, look at the insulin resistance in this group because you just cannot do that. We treated these men for three months with Sustanon, which is testosterone esters 200 mg every fortnight with a one-month washout between cross-over. The mean age was 64 years, the mean BMI was 33, waist circumference 115.1. Again, the HbA1c was similar to our large population, with a mean of 7.2%, so a well-controlled population. The mean total testosterone was 8.63 ranging from 2.34 to 11.62. Mean SHBG 27.4 and with insulin resistance you expect SHBG to be low, but we had a wide range from 11.7 to 63.4. The mean bio-available testosterone in the population was 2.73, again at the limit of quite significant hypogonadism.
We measured the HOMA and the HbA1c lipid and other estimations but I am not going to concentrate on other things. We did the waist circumference, the waist-hip ratio, which is quite good if you have the same men and you are looking at the same study, and the percentage body fat.
We showed in this study that the HOMA actually reduced on testosterone significantly.
The HbA1c also fell; the waist circumference fell as did the waist-hip ratio.
I am just going to put up a slide with all the data, which again is complicated but these are the parameters we looked at. I just want to point out in the bold here what was significant. I said the HOMA and the HbA1c was significant, and the mean HbA1c reduction was 0.37%, which is consistent with some other drugs over this period of time that are commonly used in the management of diabetes. The fasting blood glucose fell by 1.53 nmol/L. Interestingly, the cholesterol fell 0.4 nmol/L. There was no effect on HDL, LDL, trigylcerides, waist-hip ratios and waist circumference reduced but no changes in percentage body fat or BMI and fat-free mass, and the blood pressure was unaltered.
We have initially shown, it’s a cross-over study so it’s not a long-term study, we have shown over a short period of time you can improve the insulin resistance glycaemic control and this appears to be linked with the reduction in waist-circumference. What is the mechanism of this? What is the potential mechanism?
Well, we know visceral obesity is very strongly associated with insulin resistance. This is a study from 1997; this is the glucose disposal against the surface area of visceral tissue volume. So the more adipose sites you have, the greater the resistance, whereas, if you increase the amount of total subcutaneous adipose tissue, there is no increase in the insulin resistance. I think this is an important slide.
How can we explain it? The hypogonadal visceral obesity cycle was put forward by Cohen a few yeas ago, but I have developed it a bit. So you have the number of adipose sites, the higher the adipose site population, the greater the insulin resistance. However, with more adipose sites you have more of this enzyme aromatase, which breaks down testosterone to oestrogen, which will lead to lower testosterone levels. The lower testosterone levels will be compounded by the effects of leptin resistance on the pituitary gland, and also the reduced LH pulse amplitude, which is seen in obesity. This leads to an increased activity of the hormone sensitive lipoprotein lipase and then that increases triglyceride uptake and increases the number of adipose sites. This leads to this vicious circle of increasing adipose site number reducing testosterone and so forth and so on.
Just coming back to when I mentioned tumour necrosis factor earlier. We published this in JCEM in 2004. We took hypogonadal men, these were hypogonadal men with coronary heart disease, not diabetes, and we looked at the change in cytokines. Essentially the biggest fall in cytokines, this is on treatment and this is with placebo, was with the tumour necrosis factor. There were little changes with IL-1, none with IL-6, and an improvement of IL-10. But is it TNF comes in quite high concentration from visceral fat, so what are we seeing here, is it the reduction in visceral fat due to testosterone, which is leading to the lower TNF? And TNF is a very proatherogenic cytokine.
So, potentially modifiable cardiovascular risk factors by testosterone are visceral obesity, insulin resistance linked to diabetes, and hypercholesterolaemia. Other studies have shown longer-term benefits of testosterone are hypertension and coagulative state, but also there is this inflammatory markers, which will be a benefit, which may all come together to reduce the risk of heart attacks and strokes.
I just want to draw the UKPDS into this because the risk reduction associated with one percent lowering of HbA1c produces amazing effects 21% relative risk reduction in any diabetes end point 37% reduction in microvascular complications, 14% MI, 19% cataract extraction, 16% heart failure, and amputation or death due to peripheral vascular disease 43%. So, if we can do anything to try and reduce the Hb1Ac by picking these patients out and giving them testosterone replacement, we are adding to their improvement. Clearly longer-term studies are needed.
There are areas of testosterone replacement which I haven’t discussed with diabetes but we have been doing work, as you may know, before with testosterone improving cardio ischaemia with angina. Testosterone we have also shown published in European Heart Journal last month that over 12 months we improved functional capacity, and also New York Heart Association classing people with moderate heart failure. A lot of diabetics die in heart failure. There is also the testosterone replacement improvement response to PDE-5 inhibitors. The question is how is testosterone doing all of these things? It’s got its action through the classic androgen receptor, but there are also non-genomic effects, and the effects on blood vessels, which shown is actually acting as a rapid vasodilator through the L-calcium channel blocker by acting the side the nifedipine binds. So there is potential to use testosterone in the diabetic population appropriately in several patients.
So, in summary, low testosterone levels in men with type II diabetes are quite high. There is an association of low testosterone with visceral obesity and testosterone replacement leads to a reduction in visceral obesity and insulin resistance with an improvement in glycaemic control. Any factor which reduces insulin resistance should reduce the overall cardiovascular risk. Clearly what we need is larger long-term studies in the diabetic population to actually establish whether it reduces outcome.
I just want to thank D. H. Kapoor here, who has done a lot of work on diabetes and we are quite a group of cardiologists and endocrinologists working on this area.
Thank you.
Copyright © 2006 ISSAM