| Chapter one:A Meta-Analysis of Erectile Dysfunction and Risk of Clinical Cardiovascular Events Base on Common PopulationBackground:Erectile dysfunction (ED) is a multifactorial disease of the aging male affecting millions of men worldwide. The prevalence of ED increases with age: it affects 5% to 10% of men aged 40 years and 40% to 60% of men aged 70 years. From the mid 1980s to the mid 1990s in 38 populations in 21 countries worldwide, the WHO Monitoring trends and determinants in cardiovascular disease (MONICA) project had registered 166,000 myocardial infarction (MI) patients.ED and CVD are frequent co-morbid conditions and share the same risk factors including cigarette smoking, dyslipidaemia, elevated low-density lipoprotein cholesterol and low high-density lipoprotein cholesterol levels, hypertension and type 2 diabetes mellitus. The same vascular/endothelial injuries that occur in the cardiovascular are likely to occur in the cavernosal arteries, the primary arteries supplying penile erectile tissue.ED has long been considered as a secondary complication of CVD or regarded as a late consequence of generalized arterial disease. But now A body of evidence from basic science and clinical research is emerging to provide a compelling argument that ED is potentially an independent marker of subsequent CVD.Meta-analyses allow for the pooling and quantification of results from different studies. So far, meta-analyses comparing the risk of CVD events in subjects with ED and the subjects without have not been performed. Moreover, different studies have used different measurement methods, studied different populations and inconsistent results have been reported. Therefore, we performed a meta-analysis to provide an overview of the relevant studies and gain a more robust estimates of the predictive value of ED for clinical CVD events.Method:A literature search was done to identify all relevant studies that compared ED subjects and health subjects for the risk of CVD events. The Medline, Embase and Cochrane Library were searched systematically for all articles published between 1 January 1966 to 30 September 2009, without any restriction, which included the following terms in their titles, abstracts or keyword lists:erectile dysfunction (e.g., impotence, sexual dysfunction), cardiovascular disease (e.g., coronary heart disease, cerebrovascular disease, peripheral vessels disease). This search strategy was performed iteratively until no new relevant article was found.Studies were included if they agreed with the prior defined criteria No established quality assessment guidelines were available on cohort studies, and the assessment instrument designed for observational study by Liu et al was adapted for this analysis. He designed the assessment instrument according to a critical review checklist of the Dutch Cochrane Centre which was proposed by MOOSE.Data were abstracted independently by two reviewers and reached consensus on all details. We extracted the following information from each article:authors, published year, country, age range, study design, smoking rate, diabetic rate, cardiovascular events, participant number, adjusting variables, duration of follow-up and ED definition. When available, adjusted RR estimates and corresponding 95% CI were extracted and used in the meta-analysis.For the meta-analyses, both the fixed-effects model (weighted with inverse variance) and the random-effects model were considered according heterogeneity. Pooled estimates of efficacy were calculated using the Mantel-Haenszel fixed-effects model first. But if there was heterogeneity; the following methods were used to explored the source of heterogeneity:(a) subgroup analysis; (b) sensitivity analysis performed by excluding the trials which potentially biased the results. If the heterogeneity still existed, the DerSimonian and Lair random-effects model was used. For each meta-analysis, we assessed the between-study heterogeneity using the X2 and the I2 statistics, which assess the appropriateness of pooling the individual study results. A P-value of< 0.10 was interpreted as evidence of greater heterogeneity among the combined studies than what would be expected by chance alone.We used the Begg's and Egger's tests to evaluate the presence of publication bias regarding our primary end points. P< 0.05 indicated bias, and P> 0.05 indicated no publication bias. All statistical calculations were performed by Stata version 9.2 (Stata Corporation, College Station, Texas, USA).Results. The search strategy generated 2,291 references:MEDLINE (N= 2042), EMBASE (n-191), Cochrane Library (N-58), A total of 137 potentially eligible studies were identified by literature search. We excluded 75 studies that were not report the outcome of CVD event,29 were just reviews and 12 were editor articles.7 were excluded because they were not cohort design study,4 were base on diabetic population,2 paper failed to report enough data. Finally, we identified eight full-text articles that met the inclusion criteria. Of the eight articles, two reported findings from the same participants. Thus, data from these two publications were combined, giving a total of seven cohort studies that were included in the analysis.Cardiovascular disease events:The relationship between ED and the risk of CVD events was explored in the seven cohort studies. Of the six studies which reported a positive association between ED and CVD events, only one had no statistically significant. One study reported that ED was not a risk factor for CVD event. Pooled analyses of RR in fixed-effects model showed that subjects with ED had a higher overall adjusted risk of CVD events comparing with the healthy subjects and there were no heterogeneity among the studies (RR 1.47,95% CI,95% CI, 1.29-1.66, P<0.001; P for heterogeneity=0.152; I2= 36.2%).Cardiovascular diseases:Of the seven cohort studies, four reported CVD as endpoint, one reported CAD, one reported CVD events; two articles reported PVD and AMI from the same study. All the four studies which reported CVD as endpoints recorded a higher rate of CVD in the ED subjects, only one of them was not statisticlly significant. After pooling all available data for the 16,044 subjects, the rate of CVD was significantly higher in the ED group and there was no heterogeneity among the studies (RR 1.4195% CI,1.22-1.64, P<0.001; P for heterogeneity= 0.289; I2= 20.1%).All-cause mortality:Two studies recorded a higher rate of all-cause mortality in the ED group, although this was statistically significant in only one study. The pooled ratio of RR for all-cause mortality from all the two studies was significantly higher in the ED group and there was no heterogeneity among the studies (RR 1.23, 95%CI,1.02-1.48, P= 0.034; P for heterogeneity= 0.648; I2= 0.0%).Myocardial infarction:There was a higher rate of MI in the ED group in all two studies, but only one study reached statistically significant. As compared to healthy subjects, those with ED had a higher overall adjusted RR of MI and there was no heterogeneity among the studies (RR of 1.43,95% CI,1.10-1.85, P= 0.007; P for heterogeneity= 0.162; I2= 48.8%).CVD in Follow-up:Seven cohort studies had an follow-up ranging from 1 year to 15 years. The overall adjusted RR was decreasing with the time of follow-up. The pooled RRs were 1.63 (95% CI,1.43-2.21; P< 0.001; P for Heterogeneity= 0.186; I2= 37.7%) at 1 to 7 years [30,31,32,35,37] follow up and 1.37 (95% CI,1.22-1.64; P < 0.001; P for Heterogeneity= 0.252; I2= 27.4%) at≥7 years[33,34,36].Publication Bias:Begger's funnelplot and Egger's test were conducted to assess the publication bias of literatures. The shape of funnel plots did not reveal any evidence of funnel plot symmetry. The statistical results still did not show publication bias(Begg's test P= 1.000; Egger's test P= 0.622).Conclusion. The overall adjusted RR indicated that the subject with ED have an 47% inceased RR for developing CVD events and 41% for CVDcompare with the healthy subjects. The Massachusetts Male Aging Study (MMAS) reported a similar result.The overall adjusted RR of developing all-cause mortality (RR=1.23) was similar to MMAS (RR=1.26).There is evidence of an increased risk of CVD events for patients with ED. The ED patients are advised to examine for systemic vascular disease. However, to help break geographical limitation and potentially confounded association, multicentres and better-controlled studies are required. PED5i may reduce the morbidity and mortality of caridovacular events, further long-term follow-up study should be conducted. ED provides us with a window of opportunity for preventing or modifying vascular disease. We need to look through the window and seize the opportunity.Chapter two:A Meta-Analysis of Erectile Dysfunction and Risk of Clinical Cardiovascular Events Base on Diabetic PopulationBackground:Among subjects with diabetes mellitus, ED is particularly common. Approximately 25-85% of diabetic patients experience ED. Diabetic men have increased prevalence, severity, and rate of progression of ED relative to nondiabetic men. A large Italian diabetic population found that the prevalence of ED was 26% among patients with Type 1 diabetes and 37% among those with Type 2 diabetesED may precede CHD in diabetic men, probably because DM first involves endothelial dysfunction and atherosclerosis in small penile arteries, and as time passes it might cause atherosclerosis in larger arteries such as coronary arteries. Therefore, it is possible that ED in adult diabetic men may precede CHD.As we had described in chapter one. In the common population, the ED patients have a higher risk of subsequent cardiovascular events than the health people. ED can be considered as an independent risk factor for cardiovascular events. Whereas in the high risk population of cardiovascular events, such as diabetes population, the appearance of ED could be considered as a stronger early sign of cardiovascular events. So, we conduct a meta analysis in the diabetic population, the risk of cardiovascular events in the patient with ED compare to the ones without.Method:The Medline, Embase and Cochrane Library were searched systematically for all articles published between 1 January 1966 to 30 September 2009, without any restriction, which included the following terms in their titles, abstracts or keyword lists:erectile dysfunction (e.g., impotence, sexual dysfunction), cardiovascular disease (e.g., coronary heart disease, cerebrovascular disease, peripheral vessels disease). This search strategy was performed iteratively until no new relevant article was found.Studies were included if they agreed with the prior defined criteria. Data were abstracted independently by two reviewers and reached consensus on all details. We extracted the following information from each article:authors, published year, country, age range, study design, cardiovascular events, participant number, adjusting variables, duration of follow-up and ED definition. When available, adjusted RR estimates and corresponding 95% CI were extracted and used in the meta-analysis.For the meta-analyses, both the fixed-effects model (weighted with inverse variance) and the random-effects model were considered according heterogeneity Pooled estimates of efficacy were calculated using the Mantel-Haenszel fixed-effects model first. But if there was heterogeneity; the following methods were used to explored the source of heterogeneity:(a) subgroup analysis; (b) sensitivity analysis performed by excluding the trials which potentially biased the results. If the heterogeneity still existed, the DerSimonian and Lair random-effects model was used. For each meta-analysis, we assessed the between-study heterogeneity using the X2 and the I2 statistics, which assess the appropriateness of pooling the individual study results. A P-value of< 0.10 was interpreted as evidence of greater heterogeneity among the combined studies than what would be expected by chance alone.We used the Begg's and Egger's tests to evaluate the presence of publication bias regarding our primary end points. P< 0.05 indicated bias, and P> 0.05 indicated no publication bias. All statistical calculations were performed by Stata version 9.2 (Stata Corporation, College Station, Texas, USA).Results. The search strategy generated 1,113 references:MEDLINE (N= 1040), EMBASE (n= 72), Cochrane Library (N= 1), A total of 27 potentially eligible studies were identified by literature search. We excluded 10 studies because of overlap,5 studies didn't clearly define the relations between ED and cardiovascular events,2 were just reviews and 1 was editor articles,5 paper failed to report enough data. Finally, we identified 4 full-text articles that met the inclusion criteria.Cardiovascular disease events:The relationship between ED and the risk of CVD events was explored in the four studies. They all reported a positive association between ED and CVD events in diabetic patients and all have statistic significance. Pooled analyses of RR in random-effects model showed that subjects with ED had a higher overall adjusted risk of CVD events comparing with the subjects without and there were heterogeneity among the studies (RR,1.92; 95% CI,1.19-3.10, P for heterogeneity< 0.001, I2= 90.4%).Cohort studies:Because there are heterogeneity among the studies, we investigate the source of the heterogeneity by pooled the RRs excluding one study every time. Whereas, the heterogeneity still existed. So we investigate the source of the heterogeneity by subgroup analysis.There were two studies performed by cohort designed and they all reported a positive association between ED and CVD events in diabetic patients. Pooled analyses of RR in fixed-effects model showed that subjects with ED had a higher overall adjusted risk of CVD events comparing with the subjects without and there were no heterogeneity among the studies RR,1.21; 95% CI,1.10-1.34, P for heterogeneity= 0.155, I2= 50.5%)Coronary heart disease (CHD):Of the four cohort studies, three reported CHD as endpoint, one reported CAD. All the three studies recorded a higher rate of CHD in the ED subjects and have statistic significance. After pooling all available data for the 8,870 subjects, the rate of CHD was significantly higher in the ED group and there was no heterogeneity among the studies (RR,2.02; 95% CI,1.10-3.71, P for heterogeneity= 0.002,12= 83.7%)Publication Bias:Begger's funnelplot and Egger's test were conducted to assess the publication bias of literatures. The shape of funnel plots did not reveal any evidence of funnel plot symmetry. The statistical results still did not show publication bias(Begg's test P= 0.724; Egger's test P= 0.118).Conclusion. The overall adjusted RR indicated that the subject with ED have an twice increasing RR for developing CVD events compare with the subjects without, and so the CHD. The Gazzaruso et al reported a similar result.This chapter meta analysis showed that diabetes patients with ED have a higher risk of cardiovascular events. Due to longer intervals time between the appearance of ED and subsequence of cardiovascular events, ED can be used as a earlier signal of cardiovascular events. The diabetic people with ED should also examine cardiovascular disease and potential systemic vascular diseases to make early prevention and change bad habits. Similarly, to help break geographical limitation and potentially confounded association, multicentres and better-controlled studies are required. |
PDF Full Text Request