Hamling J, Lee P, Weitkunat R, Ambuhl MFacilitating meta-analyses by deriving relative effect and precision estimates for alternative comparisons from a set of estimates presented by exposure level or disease category. Stat Med 27: 954-970
P.N. Lee Statistics and Computing Ltd., 17 Cedar Road, Sutton, Surrey, U.K. Statistics in Medicine
(Impact Factor: 1.83).
03/2008; 27(7):954-70. DOI: 10.1002/sim.3013
Epidemiological studies relating a particular exposure to a specified disease may present their results in a variety of ways. Often, results are presented as estimated odds ratios (or relative risks) and confidence intervals (CIs) for a number of categories of exposure, for example, by duration or level of exposure, compared with a single reference category, often the unexposed. For systematic literature review, and particularly meta-analysis, estimates for an alternative comparison of the categories, such as any exposure versus none, may be required. Obtaining these alternative comparisons is not straightforward, as the initial set of estimates is correlated. This paper describes a method for estimating these alternative comparisons based on the ideas originally put forward by Greenland and Longnecker, and provides implementations of the method, developed using Microsoft Excel and SAS. Examples of the method based on studies of smoking and cancer are given. The method also deals with results given by categories of disease (such as histological types of a cancer). The method allows the use of a more consistent comparison when summarizing published evidence, thus potentially improving the reliability of a meta-analysis.
Available from: Sudong Liang
- "P < 0.05 (twosided ) was considered statistically significant. When a study reported risk estimates and 95 % CIs relative to a reference category other than the lowest height, we recalculated the RRs using the lowest one as reference by the method proposed by Hamling et al. (2008). In the quantitative analyses, the results for the association between height and kidney cancer risk are described continuously (per 10-cm increase in height). "
[Show abstract] [Hide abstract]
The aim of this study was to perform a meta-analysis to summarize the available evidence from prospective studies on the association between height and kidney cancer risk.
Relevant studies were identified through the MEDLINE and EMBASE databases up to July 2014, as well as through the references from the retrieved articles. Relative risks (RRs) from individual studies were pooled by using a random-effects model.
A total of fourteen prospective studies of adult height and kidney cancer risk with 18,766 cases were included in the meta-analysis. Overall, per 10-cm increase in height was associated with an increased risk of kidney cancer (RR 1.23; 95 % confidence interval 1.18-1.28, I (2) = 11.8 %). Subgroup analysis showed a basically consistent result with the overall analysis. There was no evidence of publication bias.
High adult height was positively associated with the risk of kidney cancer in both men and women in this meta-analysis. Future prospective studies are needed to determine the generalizability of these findings to non-Caucasians.
Journal of Cancer Research and Clinical Oncology 11/2014; 141(10). DOI:10.1007/s00432-014-1870-5 · 3.08 Impact Factor
Available from: PubMed Central
- "When possible, nondrinkers were chosed as the reference category; however, in several studies, occasional drinkers were included in the reference category. When more than one estimate in a study fell within the range considered for moderate or heavy alcohol consumption, the corresponding estimates were pooled using the Hamling et al.  method, thus taking into account their correlation. "
[Show abstract] [Hide abstract]
Epidemiologic studies assessing the relationship between alcohol consumption and the risk of age-related cataracts (ARCs) led to inconsistent results. This meta-analysis was performed to fill this gap.
Eligible studies were identified via computer searches and reviewing the reference lists of these obtained articles. Pooled estimates of the relative risks (RR) and the corresponding 95% confidence Intervals (CI) were calculated using random effects models.
Seven prospective cohort studies involving a total of 119,706 participants were ultimately included in this meta-analysis. Pooled results showed that there is no substantial overall increased risk of ARC due to heavy alcohol consumption. The estimated RRs comparing heavy drinkers versus non-drinkers were 1.25 (95% CI: 1.00, 1.56) for cataract sugery, 1.06 (95% CI: 0.63, 1.81) for cortical cataracts, 1.26 (95% CI: 0.93, 1.73) for nuclear cataracts, and 0.91 (95% CI: 0.32, 2.61) for posterior subcapsular cataracts (PSCs), respectively. No significant associations between moderate alcohol consumption and cataracts were observed. The pooled RRs comparing moderate drinkers versus non-drinkers were 0.90 (95% CI: 0.64, 1.26) for cataract surgery, 0.97 (95% CI: 0.75, 1.25) for cortical cataracts, 0.91 (95% CI: 0.76, 1.08) for nuclear cataracts, and 0.97 (95% CI: 0.49, 1.91) for PSCs, respectively.
This meta-analysis suggests that there is no substantial overall increased risk of ARC due to alcohol intake. Because of the limited number of studies, the findings from our study must be confirmed in future research via well-designed cohort or intervention studies.
PLoS ONE 09/2014; 9(9):e107820. DOI:10.1371/journal.pone.0107820 · 3.23 Impact Factor
Available from: Matteo Rota
- "We defined the amount of alcohol consumption as nondrinkers , light ( 1 drink, or 12.5 g of ethanol per day), moderate to heavy drinkers (>1 drink, or >12.5 g of ethanol per day). When more than one category of alcohol consumption fell in the same level, we combined the corresponding estimates using the method proposed by Hamling et al. , that allows to derive a set of pseudo-numbers of cases and non-cases that can be combined to calculate adjusted risk estimates, taking correlation into account. We used random-effects models to derive pooled meta-analytic estimates . "
[Show abstract] [Hide abstract]
ABSTRACT: The association between alcohol and leukemia risk has been addressed in several studies in the past two decades, but results have been inconsistent. Therefore, we conducted a systematic review and meta-analysis to quantify the dose–risk relation. Through the literature search up to August 2013, we identified 18 studies, 10 case-control and 8 cohorts, carried out in a total of 7142 leukemia cases. We derived pooled meta-analytic estimates using random-effects models, taking into account the correlation between estimates, and we performed a dose–risk analysis using a class of nonlinear random-effects meta-regression models. Stratified analyses were carried out on leukemia subtypes and groups, in order to identify possible etiologic differences. Compared with nondrinkers, the relative risks (RRs) for all leukemia were 0.94 [95% confidence interval (CI), 0.85–1.03], 0.90 (95% CI, 0.80–1.01) and 0.91 (95% CI, 0.81–1.02) for any, light (≤1 drink/day) and moderate to heavy (>1 drink/day) alcohol drinking, respectively. The summary RRs for any alcohol drinking were 1.47 (95% CI, 0.47–4.62) for acute lymphoblastic leukemia, 0.94 (95% CI 0.77–1.15) for chronic lymphocytic leukemia, 1.02 (95% CI, 0.86–1.21) for acute myeloid leukemia and 0.93 (95% CI 0.75–1.14) for chronic myeloid leukemia. The subgroup analysis on geographical area for all leukemia combined showed RRs of 0.84 (95% CI, 0.76–0.93), 0.92 (95% CI, 0.83–1.01) and 1.32 (95% CI, 1.02–1.70) for studies conducted in America, Europe and Asia, respectively. We did not find an increased risk of leukemia among alcohol drinkers. If any, a modest favorable effect emerged for light alcohol drinking, with a model-based risk reduction of approximately 10% in regular drinkers.
Cancer Epidemiology 06/2014; 38(4):339-345. DOI:10.1016/j.canep.2014.06.001 · 2.71 Impact Factor
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.