Performance of variable selection methods for assessing the health effects of correlated exposures in case-control studies

Objectives

There is growing recognition that simultaneously assessing multiple exposures may reduce false positive discoveries and improve epidemiological effect estimates. We evaluated the performance of statistical methods for identifying exposure–outcome associations across various data structures typical of environmental and occupational epidemiology analyses.

Methods

We simulated a case–control study, generating 100 data sets for each of 270 different simulation scenarios; varying the number of exposure variables, the correlation between exposures, sample size, the number of effective exposures and the magnitude of effect estimates. We compared conventional analytical approaches, that is, univariable (with and without multiplicity adjustment), multivariable and stepwise logistic regression, with variable selection methods: sparse partial least squares discriminant analysis, boosting, and frequentist and Bayesian penalised regression approaches.

Results

The variable selection methods consistently yielded more precise effect estimates and generally improved selection accuracy compared with conventional logistic regression methods, especially for scenarios with higher correlation levels. Penalised lasso and elastic net regression both seemed to perform particularly well, specifically when statistical inference based on a balanced weighting of high sensitivity and a low proportion of false discoveries is sought.

Conclusions

In this extensive simulation study with multicollinear data, we found that most variable selection methods consistently outperformed conventional approaches, and demonstrated how performance is influenced by the structure of the data and underlying model.

http://ift.tt/2fliM35

The INTEROCC case-control study: risk of meningioma and occupational exposure to selected combustion products, dusts and other chemical agents

Background

Little is known about occupational risk factors for meningioma.

Objectives

To study whether risk of meningioma is associated with several occupational exposures, including selected combustion products, dusts and other chemical agents.

Methods

The INTEROCC study was an international case-control study of brain cancer conducted in seven countries. Data collection by interview included lifetime occupational histories. A job exposure matrix was used to derive estimates of exposure for the 12 agents. ORs for ever versus never exposed and for exposure-response using duration of exposure and cumulative exposure were derived using conditional logistic regression stratified by sex, age group, country/region, adjusted for education.

Results

These analyses included 1906 cases and 5565 controls. For 11 of the 12 agents, no excess risk was found for ever exposed. For ever exposure to oil mists, an elevated OR of 1.57 (95% CI 1.10 to 2.22, 51 exposed cases) was found. Statistically significant exposure-response relationships were observed with cumulative exposure (p=0.01) and duration of exposure (p=0.04). Among women, there were also significant trends for cumulative and duration of exposure to asbestos and excesses in the highest exposure categories for formaldehyde.

Conclusions

Most agents examined did not provoke excess risks of meningioma. The main finding from this study is that it is the first study to identify a statistical association between exposure to oil mists and meningioma. This may be a chance finding or could be due to confounding with iron exposure and further research is required to understand whether the relationship is causal.

http://ift.tt/2wghJIr

Data on mesothelioma mortality: a powerful tool for preventing asbestos-related disease

Asbestos is a disaster.1 It has been responsible for over 200 000 deaths in the USA, for 400 000 deaths in Europe and continues today to cause an estimated 180 000 deaths each year worldwide.2 Asbestos is a known cause of cancer, and the International Agency for Research on Cancer considers all forms of asbestos—including chrysotile, the form in major use today—to be proven human carcinogens.3 Asbestos causes cancer of the lung, larynx and ovary as well as malignant mesothelioma. It is associated with stomach, pharyngeal and colorectal cancer. It causes asbestos-related non-malignant respiratory disease. No form of asbestos is safe and no level of exposure is safe. The WHO,4 the International Labour Organisation5 and the Collegium Ramazzini6 have called for a global ban on all production and use of asbestos.

Despite widespread knowledge of asbestos’ dangers, global production and sales remain...

http://ift.tt/2fm4uzf

IJERPH, Vol. 14, Pages 1118: Health Care Payments in Vietnam: Patients’ Quagmire of Caring for Health versus Economic Destitution

IJERPH, Vol. 14, Pages 1118: Health Care Payments in Vietnam: Patients’ Quagmire of Caring for Health versus Economic Destitution

International Journal of Environmental Research and Public Health doi: 10.3390/ijerph14101118

Authors: Andre Pekerti Quan-Hoang Vuong Tung Ho Thu-Trang Vuong

In the last three decades many developing and middle-income nations’ health care systems have been financed via out-of-pocket payments by individuals. User fees charges, however, may not be the best approach or thenmost equitable approach to finance and/or reform health services in developing nations. This study investigates the status of Vietnam’s current health system as a result of implementing user fees policies. A recent mandate by the government to increase the universal cover to 100% attempts to tackle inadequate insurance cover, one of the four major factors contributing to the high and increasing probability of destitution for Vietnamese patients (the other three being: non-residency, long stay in hospital, and high cost of treatment). Empirical results however suggest that this may be catastrophic for low-income earners: if insurance cover reimbursement decreases below 50% of actual health expenditures, the probability of Vietnamese falling into destitution will rise further. Our findings provide policy implications and directions to improve Vietnam’s health care system, in particular by ensuring the utilization of health services and financial protection for the people.

http://ift.tt/2fs2qt1

IJERPH, Vol. 14, Pages 1119: Comparison of Malaria Simulations Driven by Meteorological Observations and Reanalysis Products in Senegal

IJERPH, Vol. 14, Pages 1119: Comparison of Malaria Simulations Driven by Meteorological Observations and Reanalysis Products in Senegal

International Journal of Environmental Research and Public Health doi: 10.3390/ijerph14101119

Authors: Ibrahima Diouf Belen Rodriguez-Fonseca Abdoulaye Deme Cyril Caminade Andrew Morse Moustapha Cisse Ibrahima Sy Ibrahima Dia Volker Ermert Jacques-André Ndione Amadou Gaye

The analysis of the spatial and temporal variability of climate parameters is crucial to study the impact of climate-sensitive vector-borne diseases such as malaria. The use of malaria models is an alternative way of producing potential malaria historical data for Senegal due to the lack of reliable observations for malaria outbreaks over a long time period. Consequently, here we use the Liverpool Malaria Model (LMM), driven by different climatic datasets, in order to study and validate simulated malaria parameters over Senegal. The findings confirm that the risk of malaria transmission is mainly linked to climate variables such as rainfall and temperature as well as specific landscape characteristics. For the whole of Senegal, a lag of two months is generally observed between the peak of rainfall in August and the maximum number of reported malaria cases in October. The malaria transmission season usually takes place from September to November, corresponding to the second peak of temperature occurring in October. Observed malaria data from the Programme National de Lutte contre le Paludisme (PNLP, National Malaria control Programme in Senegal) and outputs from the meteorological data used in this study were compared. The malaria model outputs present some consistencies with observed malaria dynamics over Senegal, and further allow the exploration of simulations performed with reanalysis data sets over a longer time period. The simulated malaria risk significantly decreased during the 1970s and 1980s over Senegal. This result is consistent with the observed decrease of malaria vectors and malaria cases reported by field entomologists and clinicians in the literature. The main differences between model outputs and observations regard amplitude, but can be related not only to reanalysis deficiencies but also to other environmental and socio-economic factors that are not included in this mechanistic malaria model framework. The present study can be considered as a validation of the reliability of reanalysis to be used as inputs for the calculation of malaria parameters in the Sahel using dynamical malaria models.

http://ift.tt/2yoG1kc

IJERPH, Vol. 14, Pages 1120: Mercury, Fatty Acids Content and Lipid Quality Indexes in Muscles of Freshwater and Marine Fish on the Polish Market. Risk Assessment of Fish Consumption

IJERPH, Vol. 14, Pages 1120: Mercury, Fatty Acids Content and Lipid Quality Indexes in Muscles of Freshwater and Marine Fish on the Polish Market. Risk Assessment of Fish Consumption

International Journal of Environmental Research and Public Health doi: 10.3390/ijerph14101120

Authors: Joanna Łuczyńska Beata Paszczyk Joanna Nowosad Marek Łuczyński

Mercury content and fatty acids in muscles of Perca fluviatilis L. (European perch), Leuciscus idus L. (ide), Cyprinus carpio L. (European or common carp), Oncorhynchus mykiss Walb. (rainbow trout), Platichthys flesus L. (European flounder). and Clupea harengus L. (bream) from the Polish market were investigated. The total mercury was processed with AAS. The fatty acids were analyzed by gas chromatography. The concentration of mercury in muscles varied from 0.006 to 0.138 mg/kg and decreased as follows: perch ≈ ide > flounder > herring ≈ bream ≈ rainbow trout > carp (p ≤ 0.05). There were only significant positive correlations between body weight and mercury content in muscle tissue of carp (r = 0.878), flounder (r = 0.925) and herring (r = 0.982) (p ≤ 0.05). The atherogenic index (AI), thrombogenicity index (TI) and flesh-lipid quality index (FLQ) were calculated as follows 0.33–0.70 (IA), 0.16–0.31 (IT) and 13.01–33.22 (FLQ). Hypocholesterolemic (OFA) and hypercholesterolemic fatty acids (DFA) in muscles of fish ranged from 18.26 to 23.01 and from 73.91 to 78.46, respectively. In most cases, there were not significant correlations between size (body weight and total length) and fatty acids in the muscles of the examined fish (p > 0.05). The Target Hazard Quotient (THQ) values were below 1, which shows that there is no non-carcinogenic health risk to the consumer by consuming the examined fish.

http://ift.tt/2xvMsTv

IJERPH, Vol. 14, Pages 1113: Role of High-Resolution Chest Computed Tomography in a Child with Persistent Tachypnoea and Intercostal Retractions: A Case Report of Neuroendocrine Cell Hyperplasia

IJERPH, Vol. 14, Pages 1113: Role of High-Resolution Chest Computed Tomography in a Child with Persistent Tachypnoea and Intercostal Retractions: A Case Report of Neuroendocrine Cell Hyperplasia

International Journal of Environmental Research and Public Health doi: 10.3390/ijerph14101113

Authors: Mara Lelii Maria Patria Raffaella Pinzani Rossana Tenconi Alessandro Mori Nicola Bonelli Nicola Principi Susanna Esposito

Background: Chronic interstitial lung diseases in children (chILD) are a heterogeneous group of disorders that can represent a clinical challenge for pediatric pneumologists. Among them, neuroendocrine cell hyperplasia of infancy (NEHI) is a diffuse lung disease prevalent in the first years of life that spontaneously improves over time. The clinical presentation of NEHI is indistinguishable from other interstitial lung diseases, so a correct and non-invasive diagnosis by chest computed tomography (CT) without lung biopsy might not be simple. Case presentation: An 8-month-old male infant presented with a history of chronic tachypnoea and dyspnoea since 6 months of age. The patient was born at term, with APGAR scores of 9 and 10 at 1 and 5 min, respectively. Since his second month of life, the patient suffered from abnormal breathing, which was characterized by mild tachypnoea and costal retractions that worsened during breastfeeding, crying, and respiratory infections. Bilateral inspiratory crackles, preferential to the lung bases, without oxygen desaturation were detected. A chest X-ray showed a diffuse over-inflation of the lungs, but laboratory tests did not reveal any abnormalities. High-resolution chest CT documented patchy areas of ground-glass opacity involving the right upper lobe, middle lobe, and lingula, and showed mosaic areas of air-trapping, suggesting a diagnosis of NEHI. The infant was discharged without therapy and gradually improved over time. At 1 year of age, the patient was eupnoeic and chest auscultation had normalized. Conclusions: NEHI is an interstitial disease of infancy characterized by tachypnoea from the first months of life, with a good prognosis and for which a rational diagnostic approach is crucial for making a specific, early diagnosis. Initially, clinical suspicions can be confirmed with reasonable accuracy by a CT scan of the chest. Other more invasive and more expensive investigations should be reserved for selected cases that do not show a spontaneous, favourable clinical evolution.

http://ift.tt/2xzDVjM