There has been considerable interest during the past two decades regarding the importance of endogenous testosterone to the development of cardiovascular disease in middle-aged and elderly men (71-77). Studies evaluating the relationship between endogenous testosterone and cardiovascular morbidity and mortality in men have yielded inconclusive results. Most of these studies were hospital-based, case-control studies, in which cases were either men with acute myocardial infarction or men who had survived an infarction. Of the 31 cross-sectional studies, 19 (61%) found lower plasma total or free-testosterone levels in men with myocardial infarction or coronary artery disease compared with controls (21,78,79-95), whereas the remaining studies reported no significant difference in hormone levels between cases and controls (96-107). Cross-sectional studies have also analyzed the relationship between endogenous testosterone levels and the presence of angiographically defined coronary artery stenosis. Of these 13 studies, 9 (69%) found significantly decreased total or free testosterone in patients with coronary artery disease compared with controls (79-82,85,86,89,90,95), whereas four found no significant difference (97,99,103,107). An inverse correlation with the extent of angiographically defined coronary disease was stronger for free than for total testosterone, was independent of the possible confounding effects of age and body mass, and persisted after excluding men with diabetes in one report (92). Only a single cross-sectional study has analyzed the relationship between non-SHBG (bioavailable) testosterone levels and cardiovascular disease (95). That study reported significantly lower (~20%) age-adjusted and body mass index-adjusted levels of bioavailable but not total or free testosterone in men with coronary artery disease (>75% occlusion in >1 major coronary artery) compared to men with normal coronary angiograms (95).
Cross-sectional studies are limited in their ability to infer a causal relationship between hormone levels and cardiovascular disease. For example, hospitalization, lifestyle changes, or medication use after a cardiovascular event might have altered hormone levels in these previous case-control studies. There have been relatively few prospective epidemiologic studies of testosterone and cardiovascular disease events in men. Five prospective studies did not observe a significant relationship between total testosterone (65,108-111) free testosterone (65,108,111), or SHBG (65,109,111,112) and cardiovascular disease morbidity or mortality in middle-aged and elderly men. The number of clinical events in these studies was often small, although one report had sufficient statistical power to detect a difference in total testosterone as little as 100 ng/dL between myocardial infarction cases and controls (108). The relationship between bioavailable testosterone and clinical cardiovascular disease in men has not been evaluated in prospective cohort studies.
The basis for the discrepancies in previous case-control and prospective studies of cardiovascular disease remains unresolved. The majority of previous studies measured testosterone at a single point in time, which may not accurately reflect cumulative exposure to testosterone that may be important in atherosclerosis. Testosterone levels vary widely between men, and values for individual men are subject to diurnal variation, such that a single sample may characterize an individual rather poorly (113). Some authors have suggested collecting multiple samples for an individual to improve the accuracy of steroid hormone measurements (114). Furthermore, circulating testosterone measurements do not characterize androgen sensitivity or metabolism in the arterial wall. Such misclassification of exposure in past studies may have reduced the likelihood of detecting an association between testosterone and cardiovascular disease.
The strength of association between established risk factors and coronary artery disease in epidemiologic studies is affected by the control group (115), and controls were often poorly characterized in past studies of myocardial infarction. A more stringent angiographic control group (0-24% maximal stenosis) produced much stronger associations of established risk factors with coronary artery disease than a broadly defined control group (0-49% maximal stenosis) in one report (115). Thus, the strength of relationship between testosterone and cardiovascular disease may have been weakened in some studies by the inclusion of men with moderate, subclinical atherosclerosis in the control group. It is also possible that testosterone is more strongly related to the development and progression of atherosclerosis than to the incidence of clinical cardiovascular events (i.e., myocardial infarction or sudden coronary heart disease death).
Was this article helpful?