Could androgens or zinc underlie the role of HDL-cholesterol in cardiovascular disease : a review

Abstract

Background Over the past few years, results refuting the causal role of HDL cholesterol (HDL-c) have been reported by a number of randomized controlled trials (RCTs) testing different ways of modifying HDL-c. Results from Mendelian randomization studies showed no difference in cardiovascular disease (CVD) risk among individuals with genetically different serum levels of HDL-c. The causal role of HDL-c in CVD is thus uncertain, raising the question as to whether HDL-c is a worthwhile target of public health interventions and medical treatments. The objective of these meta-analyses is to explore whether changes in HDL-c are symptomatic of prior causes instead of being a causal factor for CVD by first identifying two possible candidates—androgens and zinc—for the investigation of associations. Experimental evidence would then be investigated for whether either of them might underlie (i.e. confound) the observed association of HDL-c with CVD risk factors.

Methods This study followed the PRISMA statement. A literature search was conducted through PUBMED, MEDLINE, EMBASE, and Cochrane Central Register of Controlled Trials. Keywords of “androgens/testosterone”, “HDL”, “high-density lipoprotein”, “lipid”, “cholesterol”, “lipoprotein”, “CVD”, “cardiovascular”, “heart”, “cardiovascular disease” were used with the search period limited to January 2000 – June 2013 with only human RCTs conducted and reported in English. For locating studies concerning the effect of zinc, the keyword “zinc” was used instead of “androgens/testosterone”.

Inclusion and exclusion criteria were applied during study screening and selection. The Cochrane Collaboration’s tool for assessing risk of bias was used for quality assessment. Heterogeneity across included studies was measured using I2 statistic and publication bias was assessed via funnel plots and the Begg’s Rank Correlation test. The “trim and fill” method was also used for the correction of funnel plot asymmetry.

The meta-analyses were performed using The Comprehensive R Archive Network Program (Version R 3.0.0), using the function “metacont” from the “meta” package, where the pooled intervention effects were displayed using forest plots, with inverse variance weighting and random effects model.

Results A total of twelve and ten RCTs were identified and included in the meta-analyses of androgens and zinc respectively. There were no consistent beneficial effects of androgens on CVD observed, as the results from CVD surrogate markers were inconclusive, despite showing significant overall reduction in HDL-c levels. However, as current findings suggest that lower HDL-c levels are associated with higher cardiovascular risk, it is possible that androgens may increase that risk by influencing HDL metabolism.

On the other hand, zinc was associated with healthier CVD profile. This supports the notion of zinc as a cardioprotective agent. Nonetheless, conclusion failed to be drawn concerning the effect of zinc on HDL-c as there were contradictory results across included studies.

Conclusion The meta-analyses suggest that androgens could be a factor which lowers HDL-c and thus increases cardiovascular risk, rather than HDL-c being the direct causative agent. This research may serve as a template for more extensive search for other potentially better candidates in this new study focus in cardiovascular epidemiology.