What is known
It is well documented that testosterone deficiency is associated with cardiovascular disease 13 and increased mortality risk.7,8 Low testosterone levels are also associated with the number and severity of components of the metabolic syndrome, and may predict the onset of diabetes even in younger men.14 In line with this, among men with type 2 diabetes low levels of free testosterone are associated with carotid atherosclerosis (measured by CIMT and plaque score) 15, and diabetic men with low testosterone levels have an almost 2-fold higher mortality rate compared to those with normal testosterone levels, 17.2% vs. 9% respectively.16
Testosterone deficiency is more common in men with comorbidities, such as diabetes. The prevalence of testosterone deficiency in primary care/screening studies ranges from 19.3% 17 to 38.7% 18. In other studies, testosterone deficiency affects 43 to 50% of diabetic men.5,6 The prevalence of low free testosterone levels (calculated) in diabetic men is even higher, close to 60%.5 This raises the possibility that low testosterone levels in diabetic men may be of greater clinical importance in these patients than non-diabetic patients.
Notably, in men with type 2 diabetes or the metabolic syndrome testosterone therapy has beneficial effects on obesity parameters (body weight, waist circumference and BMI) and metabolic parameters, such as the total: HDL cholesterol ratio, insulin levels, insulin resistance and CRP (C-Reactive Protein).19-22
While improvements in established risk factors support the efficacy of a treatment, the ultimate proof is hard clinical endpoints, such as cardiac events and mortality. Currently, only two intervention studies – one in type 2 diabetic men - have investigated the effect of testosterone therapy on mortality.16,23 Both studies showed that testosterone therapy reduced mortality by half.16,23 Specifically, mortality in diabetic men receiving testosterone treatment for a mean duration of 42 months was 8.4%, compared with 19.2% in the untreated group (P=0.002).16 However, these studies did not examine the possible impact of co-treatment with statins or PDE5Is.
What this study adds
The study by Hackett et al. expands previous research by prospectively examining the association between mortality and testosterone levels in men with type 2 diabetes, and evaluating the impact of concomitant statin or PDE5I use on testosterone treatment outcomes.
857 men with type 2 diabetes, 58 to 65 years old, were screened from five primary care practices during 2007 to 2009. Of the 857 men, 175 of 637 men with total testosterone levels ≤ 12 nmol/L (346 ng/dL) or free testosterone ≤ 0.25 nmol/L received testosterone treatment with injectable testosterone undecanoate for a mean duration of 3.8 years. PDE5I and statins were prescribed to 175 of 857 and 662 of 857 men, respectively.
In 2014, after a mean follow-up of 4 years, demographic, concomitant treatment (statin and PDE5I) and mortality, major adverse cardiac events and prostate-related events data were collected from the most recent follow-up visit.
Adverse Events and Mortality risk
It was found that there were no differences in major adverse cardiac events and prostate-related events among groups.
Mortality risk was higher in men who were not on testosterone therapy or PDE5I. When analysis was adjusted for age at initial visit, men in the Normal T/untreated and Low T/treated groups had a 38% and 62% reduced risk of mortality, compared to the Low T/untreated group.
When additionally adjusting for PDE5I and statin use, the association between testosterone therapy and mortality lost significance, but the association of age at initial visit and baseline testosterone levels remained significant.
Among the 682 men not given a PDE5I, when compared to the Low T/untreated group, mortality in the Normal T/untreated and Low T/treated groups was significantly lower, 38% an 67% respectively.
Statin treatment was not significantly associated with mortality in any analyses. Blood pressure, BMI and lipids were also not significantly associated with mortality.
Among men with Low T who were not receiving testosterone therapy, use of PDE5I was associated with a 94% reduced mortality risk, and statin treatment appeared protective. In contrast, PDE5I treatment was not associated with mortality in the 175 patients on testosterone therapy.
There were 103 deaths during the study period, primarily due to cardiovascular, respiratory and cancer causes. There were marked differences in mortality in the treated groups. Use of PDE5I reduced mortality rates to a greater degree in the Normal T and Low T/untreated groups, than in the Low T/treated group. Mortality rates were 19.74%, 13.31% and 3.64% in the Low T/untreated, Normal T and Low T/treated groups, respectively, as illustrated in figure 1.
This study concluded that testosterone therapy is independently associated with reduced mortality in men with type 2 diabetes, and that use of PDE5I use was associated with decreased mortality in all patients, suggesting it has an independent protective effect.
This is the first study that has investigated the independence of testosterone therapy, PDE5I and statin use, and indicates that testosterone therapy and PDE5I confer independent protection against mortality in men with type 2 diabetes, while statins do not.
This study confirms previous findings that testosterone therapy reduces mortality 16,23 by showing that mortality was higher in men who were not on testosterone therapy.12 A particularly noteworthy finding in this study is that hypogonadal men with type 2 diabetes who received testosterone treatment (without PDE5I or statins) had a 3.7-fold reduced mortality rate than type 2 diabetic men with normal testosterone levels (who are not on PDE5I or statins).
The markedly reduced mortality is concordant with previous studies that we have covered in previous editorials:
Testosterone replacement therapy and mortality in older men
Normalization of testosterone level is associated with reduced incidence of heart attack, stroke and mortality in men
Testosterone treatment and heart attack risk - new study shows testosterone treatment can even be beneficial
In addition, the lack of effect on prostate-related events confirms the established prostate safety of testosterone therapy. For more, see:
Incidence of prostate cancer after testosterone therapy for up to 17 years
Testosterone and prostate cancer - a paradigm shift