Occupational aluminum exposure: evidence in support of its neurobehavioral impact

Aluminum is a metal with known neurotoxic properties which are linked to encephalopathy and neurodegenerative diseases. The objectives of the current meta-analysis study were: (1) to summarize neurobehavioral data obtained by epidemiological studies in occupational settings and (2) to analyze confounding within these data. The meta-analysis was based on estimates of effect sizes. Overall effect sizes were obtained by application of a random effects model. The final sample consisted of nine studies examining 449 exposed and 315 control subjects. The mean urinary aluminum concentrations in the exposed groups ranged from 13 to 133 microg/l. Six neuropsychological tests, which yielded 10 performance variables, were analyzed. Nine overall effect sizes indicated an inferior performance for the exposed group. A significant overall effect size (d(RE)=-0.43) was obtained for the digit symbol test measuring speed-related components of cognitive and motor performance. Moreover, the individual effect sizes obtained for this test suggested an exposure-response relationship. Results obtained from either raw or adjusted mean scores revealed that confounding in the data could not be excluded. The results were compared to studies not included here due to a shortage of required data. Similarities were discussed in terms of sensitivity of the tests for detecting aluminum-related changes in brain function. There was concurring evidence from different studies that urinary Al concentrations below 135 microg/l have an impact on cognitive performance. The significant effect for the digit symbol might be related to its multifaceted character which requires functioning in different components of cognitive and motor performance. This feature could possibly turn the test into a screening instrument for neurobehavioral effects. However, additional studies are necessary to verify and to differentiate the effect of aluminum on cognitive performance. From a neuropsychological perspective, implicit and explicit memory, visuo-spatial and central odor processing should be examined. A measure of verbal intelligence should be included in order to address the influence of confounding. Internationally standardized exposure measures would enhance the comparability of studies.     

有向无环图在混杂因素识别与控制中的应用及实例分析

观察性研究是流行病学病因研究常用的研究设计,但应用观察性研究进行因果推断时,常由于未经识别、校正的混杂因素的存在,歪曲暴露因素与研究结局之间的真实因果关系。传统混杂因素判断标准在实际应用中不够直观,且有一定局限性,有时甚至出现混杂因素的误判。有向无环图(DAGs)可以直观识别观察性研究中存在的混杂因素,将复杂的因果关系可视化,判断研究中需要校正的最小校正子集,并可避免传统混杂因素判断标准的局限性,结合DAGs还可以指导混杂因素校正方法的选择,在观察性研究中因果推断具有重要指导价值,DAGs在未来的流行病学研究中将有更多的应用。     

A meta-analytical approach to neurobehavioural effects of occupational toluene exposure

Twenty-two studies investigating neurobehavioural effects of toluene were reviewed. Repeatedly applied neuropsychological performance tests and appropriately documented results allowed to include 10 of the studies into a meta-analysis based on effect sizes. Mean exposure level of the studies was 57 ppm (range 20–117 ppm) toluene.

Five of the six analyses of neuropsychological tests obtained effect sizes suggesting a negative impact of toluene, but for none of the analyses a significant effect size was estimated at averaged exposure levels between 33 and 89 ppm. The tests represented the psychological domains of attention and constructional performance. Additionally, the relation between exposure conditions (level of exposure, length of exposure), potential confounders (age, verbal intelligence), and effect sizes was analysed. Only pre-exposure intellectual capacity showed a consistent relation to effect sizes.

The study suggests that homogenisation of study groups with respect to intelligence, cultural background and practice trials is important when investigating low-level exposure. Efforts to homogenize studies should be completed by a thorough documentation of feasible influences.     

Propensity score model overfitting led to inflated variance of estimated odds ratios

ObjectiveSimulation studies suggest that the ratio of the number of events to the number of estimated parameters in a logistic regression model should be not less than 10 or 20 to 1 to achieve reliable effect estimates. Applications of propensity score approaches for confounding control in practice, however, do often not consider these recommendations.Study Design and SettingWe conducted extensive Monte Carlo and plasmode simulation studies to investigate the impact of propensity score model overfitting on the performance in estimating conditional and marginal odds ratios using different established propensity score inference approaches. We assessed estimate accuracy and precision as well as associated type I error and type II error rates in testing the null hypothesis of no exposure effect.ResultsFor all inference approaches considered, our simulation study revealed considerably inflated standard errors of effect estimates when using overfitted propensity score models. Overfitting did not considerably affect type I error rates for most inference approaches. However, because of residual confounding, estimation performance and type I error probabilities were unsatisfactory when using propensity score quintile adjustment.ConclusionOverfitting of propensity score models should be avoided to obtain reliable estimates of treatment or exposure effects in individual studies.     

如何控制观察性疗效比较研究中的混杂因素：（三）混杂因素控制的敏感性分析方法

观察性疗效比较研究中混杂在所难免,在利用一些统计分析方法对已测量或未测量混杂因素加以控制后,是否消除了混杂的影响不得而知,此时需进行敏感性分析。本文介绍混杂因素处理中的敏感性分析方法。基于不同的研究,敏感性分析思路各不相同,对于已测量混杂因素可采用传统的敏感性分析方法,对于未测量混杂因素目前理论相对系统的方法主要有混杂函数法、边界因子法和倾向性评分校正法,另外Monte Carlo敏感性分析和Bayes敏感性分析也是近年来备受热议的方法。当敏感性分析结果与主要分析结果一致时,无疑提高了研究结论的可靠性。     

Simultaneous measurement of salivary cortisol and alpha-amylase: Application and recommendations

Salivary cortisol (sCort) and salivary alpha-amylase (sAA) constitute proxy measures of the two major stress response systems, i.e. the hypothalamic-pituitary-adrenal axis and the autonomic nervous system, respectively. Potentially confounding determinants of sCort and sAA may limit a reliable concurrent measurement of both biomarkers, if not adequately considered. We reviewed the most important determinants of sCort and sAA and provide recommendations for handling these potential confounders. We focused on a selection of confounders, resulting in an in-depth consideration of age, sex steroid-related factors, somatic health, acute medication, smoking, consumption of food and drinks, alcohol consumption, physical activity/fitness, and sleep. Our review further highlights the importance of the consideration of potential confounders for a reliable and valid simultaneous measurement of sCort and sAA. We further provide recommendations about which and how to handle relevant confounders in the study design or data analysis.     

疾病风险评分在药物流行病学研究中的应用

疾病风险评分通过平衡不同组间研究对象的基线疾病风险以控制高维数据结构中的混杂效应,从而减小暴露因素效应估计的偏倚,因此在利用医疗数据库探索药物疗效或不良反应等规律的药物流行病学研究中有重要的应用价值。尽管疾病风险评分方法在很多情况下具有与倾向性评分相似的作用,而且在一些特殊暴露条件的研究中具有倾向性评分与传统混杂控制方法不可比拟的优势,但目前疾病风险评分在药物流行病学研究中应用范围远不及倾向性评分广泛。基于对疾病风险评分方法在药物流行病学研究中应用价值的考量,本文阐述了疾病风险评分的原理、模型构建、评分估计和应用的方法,以期为疾病风险评分方法在药物流行病学研究中的应用提供参考。