[Basics of using the DAG programs]. / Grundlagen der Anwendung von DAG-Programmen.

The application of causal graphs is a useful tool to visualize the relationship between variables and allows the identification of causal and non-causal effects. If the strict rules of the DAG theory are followed, then it is possible to identify confounding and other sources of bias. In this article we show the backtracking algorithm to find all paths of a directed acyclic graph (DAG). The knowledge of the paths can be used to identify systematically all minimally sufficient adjustment sets. The search follows formal rules and can be done by a computer program. The adjacency list and adjacency matrix, which can be used as input for a computer program, are 2 representational forms of a causal graph.

[Directed acyclic graphs (DAGs) - the application of causal diagrams in epidemiology]. / Directed Acyclic Graphs (DAGs) - Die Anwendung kausaler Graphen in der Epidemiologie.

Causal graphs such as directed acyclic graphs (DAGs) are a novel approach in epidemiology to conceptualize confounding and other sources of bias. DAGs visually encode the causal relations based on a priori knowledge among the exposure of interest and the outcome while considering several covariates. The application of formal rules on these diagrams enables the identification of the causal and non-causal structures in the DAG. The causal effects are of interest and require no adjustment. Whereas the non-causal effects have to be checked for confounding and for which covariates adjustment is necessary. The identification of the adjustment set depends on the causal relations among the variables. The consideration of these relations is valuable because adjusting for more variables increases the risk of introducing bias. Considering every single path of a DAG allows the systematic identification of the causal structures in the DAG, and the determination of minimally sufficient adjustment sets for estimating the causal effect of the exposure on the outcome based on the underlying DAG. The aim of this paper is to provide an introduction to the basic assumptions as well as the steps for drawing and applying a DAG.

Breakfast quality and cardiometabolic risk profiles in an upper middle-aged German population.

BACKGROUND/OBJECTIVES: Little is known about relation of overall breakfast quality with cardiometabolic risk factors. Therefore, this study aimed to explore sex-specific associations between breakfast quality and cardiometabolic risk profiles in a sample of an upper middle-aged German population. SUBJECTS/METHODS: Cardiometabolic profiles of 339 men and 329 women were cross-sectionally assessed using an overall biomarker score (BScore), glycated hemoglobin (HbA1c), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglycerides (TG), blood pressure, body mass index (BMI) and waist circumference (WC). Overall breakfast quality was assessed by using (i) an a-priori defined breakfast quality score (BQS) and (ii) data-driven breakfast patterns based on principal component analysis (PCA). Multiple linear regression models for association of breakfast quality with all outcomes were adjusted for all potential confounders including overall diet quality. RESULTS: After adjustment for all potential confounders the BQS was inversely associated with the BScore (regression beta with 95% Confidence Interval: -0.29 (052-0.06)) and HbA1c (-0.12 (-0.21, -0.04)) in men; whereas no such associations were observed in women. Four breakfast (B) patterns were identified: B-processed-food pattern, B-cereal pattern, B-high fat pattern and B-dairy & cereal pattern. The B-processed-food pattern was positively associated with HbA1c (0.09(0.01, 0.18)), BMI (0.16 (0.06, 0.26)), and WC (0.17 (0.8, 0.26)) in men, and BMI (0.13 (0.1, 0.25)) and WC (0.11(0.01.0.22)) in women. The B-cereal pattern was inversely associated with BScore (-0.23 (-0.45, -0.01)) and BMI (-0.11 (-0.20, -0.01)) in men and WC(-0.16 (-0.27, -0.05)) in women. The B-dairy & cereal pattern was also inversely associated with BScore (-0.26 (-0.48, -0.04)) in men but not in women. CONCLUSIONS: The overall breakfast quality was cross-sectionally associated with a healthier cardiometabolic profile, especially in upper-middle age men, independent of overall dietary quality. Such analyses should be supplemented by studies investigating the circadian sequence of food intake and metabolic consequences including hard disease endpoints.

The influence of adjustment for energy misreporting on relations of cake and cookie intake with cardiometabolic disease risk factors.

BACKGROUND/OBJECTIVES: Previous cohort studies elucidated unexpected inverse relations of cake and cookie (CC) consumption with chronic disease risk. We assessed CC intake in relation to cardiometabolic disease risk factors in a well-phenotyped population with emphasis on misreporting as the potential driving force behind inverse relations. SUBJECTS/METHODS: In a cross-sectional EPIC-Potsdam sub-study individual usual CC intake was modeled by combining 24 h recall and food frequency questionnaire data. Cardiometabolic risk factors were anthropometry, blood lipids, blood pressure (BP), physical activity and fitness. Analysis of covariance models adjusted for (i) age/education/lifestyle and (ii) additionally for energy misreporting (ratio of energy intake over energy expenditure) were used to compute mean values of risk factors for quartiles of CC intake. RESULTS: Adjustment for misreporting had considerable impact on relations of CC intake. Initial inverse links with anthropometry were reversed to direct associations. Misreporting adjustment also nullified inverse relations with triglycerides and with total cholesterol in women. Negligible associations with high density lipoprotein cholesterol turned inverse (men: cross-quartile difference (ΔQ4-Q1)=-1.7 mg/dl; women: ΔQ4-Q1=-3.6 mg/dl), so did fitness (men: ΔQ4-Q1=-1.2 ml/kg/min; women: ΔQ4-Q1=-0.9 ml/kg/min). Direct relations with total/low density lipoprotein cholesterol in men were not changed by misreporting (ΔQ4-Q1 max. 7.5 or 11.3 mg/dl). Reduced BP was observed in females with increased CC intake; only systolic BP remained relevant after misreporting adjustment (ΔQ4-Q1=-4.6 mmHg). CONCLUSIONS: The strong impact of energy misreporting on relations of CC intake with risk factors emphasizes a careful analysis and interpretation of nutritional data. We showed that apparent favorable relations of CC intake changed with a different model specification, highlighting proper modeling considerations when analyzing diet-disease relations.

The MSM program: web-based statistics package for estimating usual dietary intake using the Multiple Source Method.

BACKGROUND/OBJECTIVES: The Multiple Source Method (MSM) is a new statistical method for estimating usual dietary intake including episodically consumed foods on the basis of two or more short-term measurements such as 24-h dietary recalls. Optional information regarding habitual use or non-use of a food can be included as a covariate in the model estimating the intake, as well as a parameter for identifying consumers and non-consumers. The objective was to implement the MSM algorithms into an easy-to-use statistical program package. SUBJECTS/METHODS: The implementation was realized as a web-based application using the Perl application framework Catalyst. As the engine for the statistical calculations, the R system was used. To allow simultaneous use of the program by different users, a multiuser system with a resource bag pattern design was implemented. RESULTS: We established a software program that implements the algorithms of the MSM and allows interactive usage of the method, using standard web technologies. The program is hosted on a website established at the DIFE and can be accessed at https://nugo.dife.de/msm. The communication between users and the program web site is encrypted, securing transmitted data against unauthorized use. Users can interactively import several data sets, define the analysis model, review and export results and graphs. The use of the program is supported by online help and a user guide. CONCLUSIONS: The MSM website provides a program package that allows nutritional scientists to calculate usual dietary intakes by combining short-term and long-term measurements (multiple sources). It promotes simple access to the MSM to estimate usual food intake for individuals and populations.

[Course of left-ventricular contraction in left bundle-branch block and its hemodynamic effects]. / Linksventrikulärer Kontraktionsablauf beim Linksschenkelblock und seine hämodynamischen Auswirkungen.

The aim of the study was to analyse the left ventricular contraction pattern in left bundle branch block (LBBB), to create experimentally a comparable pattern in animals and to relate this to haemodynamic measurements. In 20 normal subjects and 16 patients with LBBB without coronary heart disease we performed computer-assisted segmental left ventricular wall motion analysis during various systolic periods using two-dimensional echocardiography. The normal subjects showed on average a uniform shortening of all segments in systole; in patients with LBBB, however, asynchronous contractions of various types and intensities were found. Examination of the contraction pattern of each LBBB patient within the confidence range of the normal subjects showed that in 94% there was an abnormally small shortening of one of the sectors at one time in the second part of systole, and in 74% in the region of the interventricular septum. A "septum index" showed significant differences (p less than 0.0025) between LBBB patients and normal subjects. By right ventricular stimulation of the apex (RVA) and the outflow tract (RVOT) we simulated these contraction patterns in 6 dogs. With RVA stimulation the left ventricular contraction pattern was nearly physiological, while with RVOT stimulation the septum movement was paradoxical. With RVA stimulation cardiac output measured by thermodilution was higher (3.45 vs. 3.11 l/min, p less than 0.002) and the left ventricular end-diastolic pressure lower (7.0 vs. 8.0 mm Hg, p less than 0.002) than on RVOT stimulation; aortic pressure and the first derivative of left ventricular pressure did not differ significantly.(ABSTRACT TRUNCATED AT 250 WORDS)