Overweight in classical phenylketonuria children: A retrospective cohort study.

PURPOSE: This cohort study aimed to determine the frequency of overweight and obesity in classical phenylketonuria children and to identify the possible influence of metabolic control on the BMI of the studied patients. PATIENTS AND METHODS: The study group included 63 classical phenylketonuria patients (40 girls and 23 boys; aged 5-16 years). Their z-score BMI, metabolic control, educational level of parents and socioeconomic status were determined. RESULTS: Twenty children were overweight or obese and only three were underweight. The percentages of overweight and obese children were 31.7% for the whole group, 21.7% (5 out of 23) for boys and 37.5% (15 out of 40) for girls. Overweight and obesity in these phenylketonuria patients was statistically significantly more frequent when compared to national reference studies (p = 0.0031). The five-year index of dietary control and the percentage of spikes exceeding 6 and 12 mg/dl (Spikes 6 and 12) indicated better metabolic control in the case of normal weight children than those who were overweight and obese (p < 0.049, p < 0.041 and p < 0.011, respectively). The odds ratio of being overweight or obese for those having poorer metabolic control (values higher vs lower than mean) was statistically significantly higher than for the remaining patients (for Spikes 12: 6.926 < 95%CI: 2.011-23.854 > ; p < 0.002). These results strongly suggest a link between overweight and diet non-compliance. CONCLUSIONS: Children with classical phenylketonuria presented higher odds of being overweight or obese as compared with reference national studies, with girls only having a higher frequency of overweight.

Fecal Fat Concentration and Excretion in the First 2 Years of Life: a Cross-Sectional Study.

OBJECTIVES: Data on pancreatic exocrine secretion in the youngest children are scarce. The aim of the study was to determine the range of normal values for fecal fat concentration (FFC) and fecal fat excretion (FFE) in infants and toddlers up to 2 years of age. METHODS: A total of 160 subjects aged 1 to 24 months (8 groups of 20: aged 1-3, 4-6 months, etc) were included in the study. In all children, FFC (%) and FFE (g/day) were assessed in 3-day stool collection. RESULTS: FFC correlated with age (r = -0.50, P < 0.0001). Of infants aged 1 to 3 months 65% had FFC higher than the <5% norm proposed for older children. The values of 90th/95th FFC percentile ranged from 9.7/13.0% at 1 month to 3.1/3.2% at 24 months. FFE did not differ between age groups. It was, however, higher than 4 and 3 g/day in 7.5% and 15.0% infants, respectively. The first detailed nomogram for the assessment of FFC and FFE in children aged 1 to 24 months was created. CONCLUSIONS: Healthy infants may have higher FFC and FFE than older children. We provide reference values, which should allow for both precise and facile FFC and FFE interpretation in clinical practice.

Fecal elastase-1 in healthy children up to 2 years of age: a cross-sectional study.

BACKGROUND: Fecal elastase-1 (E-1) levels in infants and young children may be expected to differ from those in adults and older children because of the immaturity of the gastrointestinal tract and the specificity of their diet. Despite the availability of data describing E-1 levels in the stools of preterm infants, older children, adults and subjects with malabsorption, there is still a lack of data regarding E-1 in healthy infants and toddlers. The aim of this cross-sectional study was to evaluate fecal E-1 concentrations in infants and children from 1 up to 24 months of age. MATERIAL AND METHODS: E-1 was measured in 160 healthy subjects aged 1-24 months (8 groups of 20: aged 1-3, 4-6 months, etc.) using an enzyme-linked immunosorbent assay (ELISA). RESULTS: Fecal E-1 concentrations ranged from 200 to 1695 µg/g of feces. No child had a fecal E-1 level below 200 µg/g of feces. Fecal E-1 concentrations did not significantly differ between age groups. However, fecal E-1 levels in the first 3 months were lower than in the second year of life (1-3 months vs 13-24 months, p=0.0230). A statistically significant correlation between the E-1 concentration and age was found (p=0.0007, r=0.2639; however, it does not affect the cut-off level of the reference values). The trend was rather exponential. Fecal E-1 values reached a plateau around the age of 6-10 months. CONCLUSIONS: Our study has shown that the fecal E-1 test can be reliably applied in infants and toddlers to confirm normal exocrine pancreatic function. However, within the first months of life fecal E-1 concentrations may be lower than later in life.

Determinants of Serum Glycerophospholipid Fatty Acids in Cystic Fibrosis.

The etiology of altered blood fatty acid (FA) composition in cystic fibrosis (CF) is understood only partially. We aimed to investigate the determinants of serum glycerophospholipids' FAs in CF with regard to the highest number of FAs and in the largest cohort to date. The study comprised 172 CF patients and 30 healthy subjects (HS). We assessed Fas' profile (gas chromatography/mass spectrometry), CF transmembrane conductance regulator (CFTR) genotype, spirometry, fecal elastase-1, body height and weight Z-scores, liver disease, diabetes and colonization by Pseudomonas aeruginosa. The amounts of saturated FAs (C14:0, C16:0) and monounsaturated FAs (C16:1n-7, C18:1n-9, C20:1n-9, C20:3n-9) were significantly higher in CF patients than in HS. C18:3n-6, C20:3n-6 and C22:4n-6 levels were also higher in CF, but C18:2n-6, C20:2n-6 and C20:4n-6, as well as C22:6n-3, were lower. In a multiple regression analysis, levels of seven FAs were predicted by various sets of factors that included age, genotype, forced expiratory volume in one second, pancreatic status and diabetes. FA composition abnormalities are highly prevalent in CF patients. They seem to be caused by both metabolic disturbances and independent clinical risk factors. Further research into the influence of CFTR mutations on fat metabolism and desaturases' activity is warranted.

[Detection and considering of nonlinear effects in regression modeling].

One of the basic problems solved in the research work is the search for causal relationships between the variables analyzed. Very important, but not the only requirement for the existence of causality is to demonstrate the occurrence of a statistically significant correlation and determine the shape of the present relationship (regression). In the vast majority of experimenters limited to the study of linear associations. The paper presents one of the simplest but effective method of enriching models with nonlinear terms.

[The estimation of causal effect in observational studies]. / Oszacowanie wielkosci efektu przyczynowego w badaniach obserwacyjnych.

One of the frequently encountered problems in the statistical analysis of the data is the correct interpretation of the effects occurring under the influence of some kind of treatment used by the researcher or appearing without its share of action. In the first case we are dealing with an experimental study, in the second with the observational study. In the experimental study, the researcher has full control over the procedure of random allocation of cases to a group subjected to a predetermined factor and a group not exposed to the action (randomization). In the case of an observational study, also called a retrospective study there are often additional factors that modify both assigned to the same groups as well as specific actions for the test results. The article presents an example of the cause of this phenomenon and selected methods of minimizing the mentioned influence.

[Do we always correctly interpret the results of statistical nonparametric tests]. / Czy zawsze prawidlowo interpretujemy wyniki statystycznych testów nieparametrycznych.

Mann-Whitney, Wilcoxon, Kruskal-Wallis and Friedman tests create a group of commonly used tests to analyze the results of clinical and laboratory data. These tests are considered to be extremely flexible and their asymptotic relative efficiency exceeds 95 percent. Compared with the corresponding parametric tests they do not require checking the fulfillment of the conditions such as the normality of data distribution, homogeneity of variance, the lack of correlation means and standard deviations, etc. They can be used both in the interval and or-dinal scales. The article presents an example Mann-Whitney test, that does not in any case the choice of these four nonparametric tests treated as a kind of gold standard leads to correct inference.

[Statistical inference in the presence of missing data]. / Wnioskowanie statystyczne w przypadku wystepowania brakujacych danych.

In experimental studies and particularly in medical research the missing data frequently ap-pear. Their existence can strongly affect the conclusions drawn. Therefore, it is important to treat them in appropriate way. This paper discusses three ways of dealing with missing values using example of simulated data and compares the results obtained in each case. It was found that a properly conducted multiple imputation technique gives results most consistent with those that would be obtained if the lack of data does not exist.

[Case-control models in experimental designs]. / Modele przypadek--kontrola w badaniach eksperymentalnych.

Case-control studies are one of the quickest and frequently least expensive studies both to design and carry out. Unfortunately, they are also the most vulnerable to possible biases. A crucial point is the selection of an appropriate control group. In the article three main schemes of such experiments were presented and compared.

[Confounding variables in observational studies and their influence on obtained results]. / Zmienne wiklajace w badaniach obserwacyjnych i ich wplyw na uzyskiwane wyniki.

In estimating the influence of one variable on another one we must ensure that observed effect is not caused by factors other than these under investigation. These last, called confounders of analyzed variables, influencing simultaneously both analyzed quantities may generate spurious non existing in reality relationships or change direction of these really existing.

[Binary hypothesis testing with ROC curves]. / Testowanie hipotez binarnych przy uzyciu krzywych ROC.

In medical research we frequently need to delimit some areas of the measured parameters to use them in diagnosing existence of some kind of abnormality. In the article the usage of ROC curves and comparison of obtained results to the outcomes of logistic analysis, decision trees and reference intervals is described.

[Methods of exact nonparametric inference for non-typical distributions of experimental data]. / Metody dokladnego wnioskowania nieparametrycznego w przypadku nietypowych rozkladów danych eksperymentalnych.

In medical research we frequently find data sets with specific structure such as small data sets, unbalanced, sparse or heavily tied. The peculiar properties of those sets influence the p-value which quantity is used in decision making process. Four examples of experimental data, for which estimation of asymptotic p-value leads to erroneous decision, are presented.

[Common errors in analysis of the relationship between variables]. / Najczestsze bledy popelniane przy badaniu zaleznosci miedzy zmiennymi.

One of the key research problems is the analysis of relationships between variables. Mathematical modeling is commonly used technique in quantification of these associations. This paper describes errors most frequently occurring in simple and complex models constructions such as assumption violation, influence of outliers, data clustering, influence of confounding variables, limited scope of model application. Mentioned problems were illustrated with appropriate examples.

[Application of virtual instrumentation technique in toxicological studies]. / Wykorzystanie techniki instrumentacji wirtualnej w badaniach toksykologicznych.

Research investigations require frequently direct connection of measuring equipment to the computer. Virtual instrumentation technique considerably facilitates programming of sophisticated acquisition-and-analysis procedures. In standard approach these two steps are performed subsequently with separate software tools. The acquired data are transfered with export / import procedures of particular program to the another one which executes next step of analysis. The described procedure is cumbersome, time consuming and may be potential source of the errors. In 1987 National Instruments Corporation introduced LabVIEW language based on the concept of graphical programming. Contrary to conventional textual languages it allows the researcher to concentrate on the resolved problem and omit all syntactical rules. Programs developed in LabVIEW are called as virtual instruments (VI) and are portable among different computer platforms as PCs, Macintoshes, Sun SPARCstations, Concurrent PowerMAX stations, HP PA/RISK workstations. This flexibility warrants that the programs prepared for one particular platform would be also appropriate to another one. In presented paper basic principles of connection of research equipment to computer systems were described.

[Use of rough sets theory in analysis of qualitative data in toxicology studies]. / Wykorzystanie teorii zbiorów przyblizonych do analizy danych jakosciowych w badaniach toksykologicznych.

In analysis of medical data a problem of simplified and not effective analysis of qualitative data is frequently encountered. This situation takes place particularly when data are measured in the weakest measuring scale--nominal scale. However this scale contains less detailed information than interval or ordinal scale, there exist a lot of different mathematical approaches to examine existing dependencies, among other causality. In this paper we present the exam-pie of usage of technique based on theory of rough sets to toxicological data gathered during mediumistic program "Throw smoking together with us" in 2000. The obtained results were compared with results estimated with alternative techniques published in [1].