Evangelia E. Ntzani

Replication validity of genetic association studies

The rapid growth of human genetics creates countless opportunities for studies of disease association. Given the number of potentially identifiable genetic markers and the multitude of clinical outcomes to which these may be linked, the testing and validation of statistical hypotheses in genetic epidemiology is a task of unprecedented scale. Meta-analysis provides a quantitative approach for combining the results of various studies on the same topic, and for estimating and explaining their diversity. Here, we have evaluated by meta-analysis 370 studies addressing 36 genetic associations for various outcomes of disease. We show that significant between-study heterogeneity (diversity) is frequent, and that the results of the first study correlate only modestly with subsequent research on the same association. The first study often suggests a stronger genetic effect than is found by subsequent studies. Both bias and genuine population diversity might explain why early association studies tend to overestimate the disease protection or predisposition conferred by a genetic polymorphism. We conclude that a systematic meta-analytic approach may assist in estimating population-wide effects of genetic risk factors in human disease.

Association between omega-3 fatty acid supplementation and risk of major cardiovascular disease events: a systematic review and meta-analysis.

CONTEXT Considerable controversy exists regarding the association of omega-3 polyunsaturated fatty acids (PUFAs) and major cardiovascular end points. OBJECTIVE To assess the role of omega-3 supplementation on major cardiovascular outcomes. DATA SOURCES MEDLINE, EMBASE, and the Cochrane Central Register of Controlled Trials through August 2012. STUDY SELECTION Randomized clinical trials evaluating the effect of omega-3 on all-cause mortality, cardiac death, sudden death, myocardial infarction, and stroke. DATA EXTRACTION Descriptive and quantitative information was extracted; absolute and relative risk (RR) estimates were synthesized under a random-effects model. Heterogeneity was assessed using the Q statistic and I2. Subgroup analyses were performed for the presence of blinding, the prevention settings, and patients with implantable cardioverter-defibrillators, and meta-regression analyses were performed for the omega-3 dose. A statistical significance threshold of .0063 was assumed after adjustment for multiple comparisons. DATA SYNTHESIS Of the 3635 citations retrieved, 20 studies of 68,680 patients were included, reporting 7044 deaths, 3993 cardiac deaths, 1150 sudden deaths, 1837 myocardial infarctions, and 1490 strokes. No statistically significant association was observed with all-cause mortality (RR, 0.96; 95% CI, 0.91 to 1.02; risk reduction [RD] -0.004, 95% CI, -0.01 to 0.02), cardiac death (RR, 0.91; 95% CI, 0.85 to 0.98; RD, -0.01; 95% CI, -0.02 to 0.00), sudden death (RR, 0.87; 95% CI, 0.75 to 1.01; RD, -0.003; 95% CI, -0.012 to 0.006), myocardial infarction (RR, 0.89; 95% CI, 0.76 to 1.04; RD, -0.002; 95% CI, -0.007 to 0.002), and stroke (RR, 1.05; 95% CI, 0.93 to 1.18; RD, 0.001; 95% CI, -0.002 to 0.004) when all supplement studies were considered. CONCLUSION Overall, omega-3 PUFA supplementation was not associated with a lower risk of all-cause mortality, cardiac death, sudden death, myocardial infarction, or stroke based on relative and absolute measures of association.

Genetic associations in large versus small studies: an empirical assessment

BACKGROUND Advances in human genetics could help us to assess prognosis on an individual basis and to optimise the management of complex diseases. However, different studies on the same genetic association sometimes have discrepant results. Our aim was to assess how often large studies arrive at different conclusions than smaller studies, and whether this situation arises more frequently when findings of first published studies disagree with those of subsequent research. METHODS We examined the results of 55 meta-analyses (579 study comparisons) of genetic associations and tested whether the magnitude of the genetic effect differs in large versus smaller studies. FINDINGS We noted significant between-study heterogeneity in 26 (47%) meta-analyses. The magnitude of the genetic effect differed significantly in large versus smaller studies in ten (18%), 20 (36%), and 21 (38%) meta-analyses with tests of rank correlation, regression on SE, and regression on inverse of variance, respectively. The largest studies generally yielded more conservative results than the complete meta-analyses, which included all studies (p=0.005). In 14 (26%) meta-analyses the proposed association was significantly stronger in the first studies than in subsequent research. Only in nine (16%) meta-analyses was the genetic association significant and replicated without hints of heterogeneity or bias. There was little concordance in first versus subsequent discrepancies, and large versus small discrepancies. INTERPRETATION Genuine heterogeneity and bias could affect the results of genetic association studies. Genetic risk factors for complex diseases should be assessed cautiously and, if possible, using large scale evidence.

Predictive ability of DNA microarrays for cancer outcomes and correlates: an empirical assessment.

BACKGROUND DNA microarrays are being used for many applications, including the prediction of cancer outcomes by simultaneous analysis of the expression of thousands of genes. We systematically assessed the predictive performance of this method for major clinical outcomes (death, metastasis, recurrence, response to therapy) and the correlation of gene profiling with other clinicopathological correlates of malignant disorders. METHODS Eligible reports retrieved from MEDLINE (1995 to April, 2003) were assessed for features of study design, reported predictive performance, and consideration of other prognostic factors. We searched for study variables that increased the chances that a significant association with a clinical outcome or correlate would be found. FINDINGS 84 eligible studies were identified, of which 30 addressed major clinical outcomes. A median of 25 (IQR 15-45) patients with cancer were included. Among the studies of major clinical outcomes, nine did cross-validation but it was complete in only two of them; six studies used independent validation of supervised predictive models. Smaller studies showed better sensitivity and specificity for clinical outcomes than larger studies. Only 11 studies addressing major clinical outcomes did subgroup or adjusted analyses for other prognostic factors. Across all 84 studies, significant associations were 3.5 (95% CI 1.5-8.0) times more likely per doubling of sample size and 9.7 (2.0-47.0) times more likely per ten-fold increase in microarray probes. INTERPRETATION DNA microarrays addressing cancer outcomes show variable prognostic performance. Larger studies with appropriate clinical design, adjustment for known predictors, and proper validation are essential for this highly promising technology.

Hardy–Weinberg equilibrium in genetic association studies: an empirical evaluation of reporting, deviations, and power

We evaluated the testing and reporting of Hardy–Weinberg equilibrium (HWE) in recent genetic association studies, detected how frequently HWE was violated and estimated the power for HWE testing in this literature. Genetic association studies published in 2002 in Nature Genetics, American Journal of Human Genetics, and American Journal of Medical Genetics were assessed. Data were analyzed on 239 biallelic associations using 154 distinct genotype distribution data sets where HWE could be tested. Any information on HWE was given only for 150 (62.8%) associations (92 (59.7%) data sets). Reanalysis of the data showed significant deviation from HWE in the disease-free controls of 20 associations (13 data sets), but only four of them (two data sets) were admitted in the published articles. Another four deviations (in two data sets) were observed in the combined sample of cases and controls of studies where both cases and controls were diseased, and none were reported in the papers. In all six tested multiallelic associations (six data sets), there was violation of HWE, but this was not admitted in the published articles. Power calculations showed that most studies conforming to HWE simply were largely underpowered to detect HWE deviation; for example, power to detect an inbreeding of magnitude F=0.10 exceeded 80% in only 11 (7%) of the data sets being tested. This empirical evidence suggests that, even in high profile genetics journals, testing and reporting for HWE is often neglected and deviations are rarely admitted in the published reports. Moreover, power is limited for HWE testing in most current genetic association studies.

Type 2 diabetes and cancer: umbrella review of meta-analyses of observational studies.

Objectives To summarise the evidence and evaluate the validity of the associations between type 2 diabetes and the risk of developing or dying from cancer. Design An umbrella review of the evidence across meta-analyses of observational studies of type 2 diabetes with risk of developing or dying from any cancer. Data sources PubMed, Embase, Cochrane database of systematic reviews, and manual screening of references. Eligibility criteria Meta-analyses or systematic reviews of observational studies in humans that examined the association between type 2 diabetes and risk of developing or dying from cancer. Results Eligible meta-analyses assessed associations between type 2 diabetes and risk of developing cancer in 20 sites and mortality for seven cancer sites. The summary random effects estimates were significant at P=0.05 in 20 meta-analyses (74%); and all reported increased risks of developing cancer for participants with versus without diabetes. Of the 27 meta-analyses, eventually only seven (26%) compiled evidence on more than 1000 cases, had significant summary associations at P≤0.001 for both random and fixed effects calculations, and had neither evidence of small study effects nor evidence for excess significance. Of those, only six (22%) did not have substantial heterogeneity (I2>75%), pertaining to associations between type 2 diabetes and risk of developing breast, cholangiocarcinoma (both intrahepatic and extrahepatic), colorectal, endometrial, and gallbladder cancer. The 95% prediction intervals excluded the null value for four of these associations (breast, intrahepatic cholangiocarcinoma, colorectal, and endometrial cancer). Conclusions Though type 2 diabetes has been extensively studied in relation to risk of developing cancer and cancer mortality and strong claims of significance exist for most of the studied associations, only a minority of these associations have robust supporting evidence without hints of bias.

Patellar resurfacing in total knee arthroplasty. A meta-analysis.

BACKGROUND Patellar resurfacing during total knee arthroplasty remains controversial. We aimed to evaluate the effectiveness of this technique through an evaluation of the current literature. METHODS We performed a meta-analysis of randomized controlled trials comparing total knee arthroplasties performed with and without patellar resurfacing. Outcomes of interest included the number of reoperations, the prevalence of postoperative anterior knee pain, and the improvement in various knee scores. RESULTS Ten trials assessing 1223 knees were eligible. The absolute risk of reoperation was reduced by 4.6% (95% confidence interval, 1.9% to 7.3%) in the patellar resurfacing arm (between-study heterogeneity, p < 0.01; I(2) = 60%), implying that one would have to resurface twenty-two patellae (95% confidence interval, fourteen to fifty-two patellae) in order to prevent one reoperation. Patellar resurfacing reduced the absolute risk of postoperative anterior knee pain by 13.8% (95% confidence interval, 6.4% to 21.2%), implying that one would have to resurface seven patellae (95% confidence interval, five to sixteen patellae) in order to prevent one case of postoperative anterior knee pain. Only four trials provided adequate data for a quantitative synthesis of the changes in the various knee scores; on the basis of those four trials, there was no difference in the mean improvement in the knee scores (standardized mean difference, 0.03; 95% confidence interval, -0.50 to 0.56). CONCLUSIONS The available evidence indicates that patellar resurfacing reduces the risks of reoperation and anterior knee pain after total knee arthroplasty. The observed effects are clinically important despite their modest magnitude. Additional, carefully designed randomized trials are required to strengthen this claim.

Impact of Schizophrenia Candidate Genes on Schizotypy and Cognitive Endophenotypes at the Population Level

BACKGROUND Aspects of cognitive function and schizotypy have been proposed as potential endophenotypes for schizophrenia. It is unknown whether the expression of these endophenotypes at the population level is modulated by the genetic variability of candidate susceptibility genes for schizophrenia. METHODS We examined the potential impact of 18 single nucleotide polymorphisms (SNPs) within the DTNBP1, NRG1, DAOA/G32, and DAAO genes, on cognition and self-rated schizotypy, in a representative population of 2243 young male military conscripts. Single SNP and haplotype associations were evaluated. RESULTS The DTNBP1 SNPs rs2619522 and rs760761 exhibited several single marker associations, the minor alleles being associated with lower attention capacity but also a decrease in positive and paranoid schizotypy scores. The DTNBP1 haplotype load had borderline associations with nonverbal IQ, paranoid schizotypy, and sustained attention. For individual NRG1 polymorphisms, isolated but weak signals of association were noted with sustained attention and working memory but not schizotypy. The risk allele of functional SNP8NRG243177 was associated with reduced spatial working memory capacity. An isolated effect of DAAO haplotype variability was noted on negative and disorganization schizotypy. No convincing association of DAOA/G32 variability was detected. CONCLUSIONS The DTNBP1 and, less so, NRG1 and DAAO variants might exert gene-specific modulating effects on schizophrenia endophenotypes at the population level.

Establishment of genetic associations for complex diseases is independent of early study findings

Numerous genetic association studies for complex diseases are performed. Investigators place emphasis on formal statistical significance (P-values <0.05), but the predictive ability of early statistically significant (‘positive’) findings is unclear. We scrutinized 55 cumulative meta-analyses of genetic associations (579 studies), in order to assess whether having statistical significance in the earliest (first) published study or in at least half among several (⩾3) early-published studies, or high statistical significance in early studies had any predictive ability for establishing or refuting the presence of the genetic association in subsequent research. In 35 associations, a first study was ‘positive’ and in 15 associations more than half of the early-published reports were ‘positive’. The average publication rate of subsequent studies increased 1.71-fold with a ‘positive’ first report. When compared against the summary results of subsequent research, sensitivity and specificity were 0.65 and 0.38 for the first reports, and 0.40 and 0.73, respectively, when at least three early studies were considered. First studies also had poor predictive ability, when we considered the estimated attributable fraction and coverage of the 95% confidence interval thereof or higher levels of statistical significance. We conclude that although ‘positive’ findings in the very first reports provide strong incentive for conducting more studies on a putative genetic epidemiological association, the statistical significance or even the magnitude of the effect of early studies cannot adequately predict eventual establishment of an association. Conversely, many genuine epidemiological associations would be missed, if research were abandoned after early underpowered ‘negative’ studies.

CYP2D6 polymorphisms and the risk of tardive dyskinesia in schizophrenia: a meta-analysis

The present study aimed to evaluate whether there is any association between CYP2D6 alleles and susceptibility to tardive dyskinesia in patients with schizophrenia under treatment. A meta-analysis considered case–control studies determining the distribution of genotypes for any CYP2D6 polymorphism in unrelated tardive dyskinesia cases and controls without tardive dyskinesia among patients with schizophrenia who were treated with antipsychotic agents. Loss of function alleles were grouped together in a single comparison, whereas other alleles (*2 and *10) were examined separately. Data were available for eight (n=569 patients), three (n=325 patients) and four (n=556) studies evaluating the effect of the loss of function alleles, the *2 allele and the *10 allele, respectively. Summary odds ratios (ORs) suggested that loss of function alleles increased the risk of tardive dyskinesia significantly [OR=1.43, 95% confidence interval (CI) 1.06–1.93, P=0.021], whereas there was no effect for *2 and inconclusive evidence for *10 (OR=0.82, 95% CI 0.50–1.32, P=0.41 and OR=1.19, 95% CI, 0.89–1.60, P=0.24, respectively). Patients who were homozygotes for loss of function alleles (poor metabolizers) had 1.64-fold greater odds of suffering tardive dyskinesia compared to other patients with schizophrenia, but the effect was not formally significant (95% CI 0.79–3.43). For the risk conferred by loss of function alleles, large studies provided more conservative estimates of a genetic effect than smaller studies (P=0.003). CYP2D6 loss of function alleles may predispose to tardive dyskinesia in patients with schizophrenia under treatment, but bias cannot be excluded.