Invited Commentary: Acne in Adolescence—Protecting the Heart but Damaging the Prostate Later in Life?

Abstract

Cutaneous manifestations of diseases have long helped observant clinicians to establish a diagnosis (1), but can cutaneous clues or other physical characteristics predict the risk of mortality or incident disease far into the future? For some time, researchers have linked with mixed success various physical manifestations to the risk of developing diseases, such as coronary heart disease. Among these physical markers are earlobe crease (2), arcus senilis (3), baldness (4), early hair graying (5), and facial wrinkling (6). Good reasons exist for studying the associations between physical markers and disease endpoints, because physical characteristics may help not only to elucidate underlying pathogenetic processes for diseases but also to detect persons at increased risk who would benefit by early intervention. In this issue of the American Journal of Epidemiology, Galobardes et al. (7) add acne vulgaris to the list of physical characteristics that may serve as risk markers for the development of disease later in life. In a retrospective follow-up study of 11,232 men who attended Glasgow University between 1948 and 1968 and whose mortality was traced into 2004, the investigators found that participants who reported having acne during adolescence had a significantly lower risk of death from coronary heart disease and an increased, although statistically not significant, risk of death from prostate cancer. For the other causes of mortality (mortality from all causes, stroke, lung cancer, colon cancer, and external causes), no significant associations were found. This study provides an excellent example of how older epidemiologic studies can be creatively used to generate new hypotheses about the potential pathogenetic pathways of various conditions. Consistent with the longstanding hypothesis that endogenous sex hormones play important roles in the pathogenesis of prostate cancer and cardiovascular disease, the authors’ view of acne largely as a manifestation of endogenous hormonal activity, principally androgen activity, has precedent. Although there are many reasons to believe that androgen activity may increase the risk of prostate cancer, epidemiologic studies have produced inconsistent results in linking circulating concentrations of androgens to the risk of prostate cancer (8–10). In the current study, the hazard ratio of 1.67 for mortality from prostate cancer was based on only 37–43 such deaths, and the confidence interval was wide and overlapped 1.0 considerably. Surprisingly, the authors highlighted the findings pertaining to prostate cancer in the title of their article as well as in the body of the text. The cardiovascular findings are perhaps more interesting and noteworthy. The authors report hazard ratios of 0.74 for mortality from total cardiovascular disease and 0.67 for coronary heart disease mortality. In both instances, the confidence interval excluded unity. Adjustment for the various covariates considered by the authors had little effect on the hazard ratios. While a series of recent reviews seem to be cautiously converging on the view that androgens may exert a favorable effect on cardiovascular disease (11–13), the roles of androgens on the cardiovascular system are likely to be complex and may be sex specific. Among men, concentrations of testosterone have been inversely related to concentrations of insulin and insulin resistance (14–21), as