Multivariate analysis of body shape in fragile X (Martin-Bell) syndrome

Human body shape measures can be more informative in studies of developmental abnormalities than distances between body landmarks. Such measures were obtained by an appropriate transformation of 34 distances between trunk/limbs and head/face landmarks in 43 men and 72 women with the Martin-Bell Syndrome (MBS), and in 99 and 103 normal men and women, respectively. The transformation of the original distances was performed by adjusting individual measures for an overall size measure using regression analysis. Thus obtained body shape variables were used in discriminant analysis in order to obtain unbiased classification probabilities of individuals having the MBS or being normal. The average percent correctly classified male and female individuals was high (93 and 87, respectively). Moreover, the body shape variables were used to obtain shape dimensions by means of principal component analysis. There was no difference between the MBS and normal individuals in the first (most important) principal component (shape dimension). This component represents the relative proportions between trunk and limb lengths and widths, or between midfacial lengths and widths. However, there were appreciable differences in some succeeding components. The problem of interpretation of shape dimensions as derived from principal component analysis and of their relevance to abnormal development in the MBS individuals is discussed.     

Cross-sectional-type weight reference values for village children under five years in Lesotho.

Cross-sectional-type reference values for weight attained are described for village children under five years in rural Lesotho (formerly Basutoland). Weight measurements derive from observations on 1317 children attending an Under-Fives clinic; it is estimated that 60-70% of the children under five in the catchment area were represented. 4585 weighings on boys and 4826 weighings for girls are included in the analysis. Figures of weight-for-age of boys and girls are given separately as centile distributions suitable for use on Growth Charts. Lesotho 50 centile approximates to 3 centile of British children and slightly exceeds 80% Harvard standard. Weight attained for age is similar, in both sexes, to reports from other less-priviledged urban and rural areas, emphasizing the relative importance of environmental as compared to genetic influences in determining weight-for-age in early childhood. It is suggested that the construction of locally-derived growth reference values is both appropriate and practicable.     

Twins: measures of birth size at different gestational ages

Means, standard deviations, and empirical centile values are given for birth weight, length and head circumference for twins born at gestational ages 31+ to 42+ weeks. The data are compared with recent results for singletons, and the twins show a deficit that becomes greater as gestational age increases. Although weight may be slightly more affected than the other two measures, the relative growth suppression for twins is quite similar in all three dimensions. Comparatively, the average size of twins born at 33+ weeks gestational age is equal to the 36th centile for singletons; but at 40+ weeks gestational age, the average size for twins is below the 5th centile for singletons.     

A comparison of US-born African-American and African-Caribbean psychiatric outpatients.

This study compares US-born African Americans with African Caribbeans attending an urban psychiatric outpatient clinic on various items pertaining to sociodemographics, psychiatric history, current psychiatric illness, and physical health. A structured chart review was performed on a sample comprised of 135 native-born African Americans and 91 African Caribbeans who had attended the clinic during an 11-year period. A total of 28 clinical variables were examined. Nine clinical variables were found to significantly differentiate the two groups in bivariate analysis, and these were entered along with three demographic control variables into a logistic regression analysis. Seven variables attained significant independent effects. African Americans were differentiated from African Caribbeans on history of greater alcoholic abuse or dependence, presence of more delusions, worse health, longer history of previous outpatient treatment, and greater clinical improvement at 6 months. African Caribbeans were found to have a greater frequency of depression and aggressivity. These data underscore the importance of examining intraracial differences in mental illness as well as pointing to the potential benefits of using intraracial comparisons to interpret interracial analyses.     

Relationships among tempo of maturation,midparent height,and growth in height of adolescent boys and girls

The influence of rate of skeletal maturation and midparent height on growth in height during adolescence was analyzed on longitudinal data for 184 boys and 166 girls from the Wrocław Growth Study. Seven biological parameters describing the shape of the growth curve in stature were derived from the Preece–Baines model 1 applied to individual serial data. Rate of skeletal maturation inferred from the difference between chronological and skeletal ages at 12 years in girls and 14 in boys. Principal components analysis of the biological parameters extracted three factors for boys and four for girls. The factors explained 90% and 97% of the total variance in boys and girls, respectively. The factors reflected attained size and the timing and intensity of the adolescent spurt in height. Multiple regression showed a high relationship between skeletal maturation rate, midparent height, and principal component scores. Three aspects of adolescent growth: size, timing, and intensity show an apparent sex difference. The timing and intensity of the spurt are highly affected by tempo of maturation in girls, but less so than in boys. It might suggest in girls a less efficient compensatory effect for reduced length of overall growth period in early maturers that increases height gain. Midparent height influences the size component during the spurt, thus adjusting the height parameters. The dependency between attained stature (height at take‐off, PHV, and adult stature), timing, and initial intensity of the growth spurt of daughters and midparent height is greater than in sons. Am. J. Hum. Biol. 13:15–22, 2001. © 2001 Wiley‐Liss, Inc.     

Sparse modeling of spatial environmental variables associated with asthma

Geographically distributed environmental factors influence the burden of diseases such as asthma. Our objective was to identify sparse environmental variables associated with asthma diagnosis gathered from a large electronic health record (EHR) dataset while controlling for spatial variation. An EHR dataset from the University of Wisconsin’s Family Medicine, Internal Medicine and Pediatrics Departments was obtained for 199,220 patients aged 5–50 years over a three-year period. Each patient’s home address was geocoded to one of 3456 geographic census block groups. Over one thousand block group variables were obtained from a commercial database. We developed a Sparse Spatial Environmental Analysis (SASEA). Using this method, the environmental variables were first dimensionally reduced with sparse principal component analysis. Logistic thin plate regression spline modeling was then used to identify block group variables associated with asthma from sparse principal components. The addresses of patients from the EHR dataset were distributed throughout the majority of Wisconsin’s geography. Logistic thin plate regression spline modeling captured spatial variation of asthma. Four sparse principal components identified via model selection consisted of food at home, dog ownership, household size, and disposable income variables. In rural areas, dog ownership and renter occupied housing units from significant sparse principal components were associated with asthma. Our main contribution is the incorporation of sparsity in spatial modeling. SASEA sequentially added sparse principal components to Logistic thin plate regression spline modeling. This method allowed association of geographically distributed environmental factors with asthma using EHR and environmental datasets. SASEA can be applied to other diseases with environmental risk factors.     

A comparison of statistical techniques for analysis of growth curves

Frequently, experiments are conducted in order to investigate the effects of various treatments on an animal's growth rate. The data from these investigations usually consist of each animal's body weight or accumulative weight gain at specific times during the experiment. The most common statistical techniques for analysis of growth rates (increments in body weight over time) consider only terminal body weights or final accumulative weight gain. In this study, we compare growth rates over the duration of the experiment and use standard simultaneous testing procedures in order to accommodate more than two treatment groups. Results obtained by comparison of regression lines randomization analysis of variance, and repeated measures analysis are presented.     

Detecting Functional Connectivity in fMRI Using PCA and Regression Analysis

A fMRI connectivity analysis approach combining principal component analysis (PCA) and regression analysis is proposed to detect functional connectivity between the brain regions. By first using PCA to identify clusters within the vectors of fMRI time series, more energy and information features in the signal can be maintained than using averaged values from brain regions of interest. Then, regression analysis can be applied to the extracted principal components in order to further investigate functional connectivity. Finally, t-test is applied and the patterns with t-values lager than a threshold are considered as functional connectivity mappings. The validity and reliability of the presented method were demonstrated with both simulated data and human fMRI data obtained during behavioral task and resting state. Compared to the conventional functional connectivity methods such as average signal based correlation analysis, independent component analysis (ICA) and PCA, the proposed method achieves competitive performance with greater accuracy and true positive rate (TPR). Furthermore, the ‘default mode’ and motor network results of resting-state fMRI data indicate that using PCA may improve upon application of existing regression analysis methods in study of human brain functional connectivity.     

Single occasion means of height and weight of Sri Lankan Tamil children from Jaffna

Little published material is available on physical growth in any Sri Lankan population. This paper reports attained mean heights and weights of middle class Tamil children from Jaffna, based on single occasion measurements of 358 boys and 517 girls ranging in age between 6 and 17 years. The cross-sectional height and weight data of both boys and girls show a typical pattern. Boys are taller and heavier up to 9 years, then girls lead in both the measurements due to their pubertal phase of rapid growth. Boys supersede them again when they enter into their rapid growth phase, i.e. in height at 13 years but their mean weights remain lower till 16 years. Thus at 17 years boys on average are 5 cm taller than girls. With reference to British standards the attained means of both boys' and girls' height remain around the 10th centile, but later shift to the 3rd centile. In weight, however, they remain below the 3rd centile and indeed much below the range during the later part of the pubertal phase. Apart from nutrition and living conditions, how much of this pattern is influenced by ethnic and tropical climatic factors is still unclear.     

Single occasion means of height and weight of Sri Lankan Tamil children from Jaffna

Little published material is available on physical growth in any Sri Lankan population. This paper reports attained mean heights and weights of middle class Tamil children from Jaffna, based on single occasion measurements of 358 boys and 517 girls ranging in age between 6 and 17 years. The cross-sectional height and weight data of both boys and girls show a typical pattern. Boys are taller and heavier up to 9 years, then girls lead in both the meassurements due to their pubertal phase of rapid growth. Boys supersede them again when they enter into their rapid growth phase, i.e. in height at 13 years but their mean weights remain lower till 16 years. Thus at 17 years boys on average are 5 cm taller than girls. With reference to British standards the attained means of both boys' and girls' height remain around the 10th centile, but later shift to the 3rd centile. In weight, however, they remain below the 3rd centile and indeed much below the range during the later part of the pubertal phase. Apart from nutrition and living conditions, how much of this pattern is influenced by ethnic and tropical climatic factors is still unclear.     

Growth curve and morphometric variables in rats: are they related?

Four growth curve variables were measured in a population of rats in order to assess their association, if any, with a suite of morphometric characters. The growth curve variables (A = asymptotic weight, R = growth rate, P = percentage of asymptotic size at inflection, and G = time to grow from 10 to 90% of asymptotic size) were derived from Richards' curves applied to longitudinal body weight data, and the morphometric characters (measured in 189-day-old rats) consisted of 4 tooth and 7 bone variables. As assessed by a canonical correlation analysis, there was a significant association between the two sets of variables, especially between A and the skeletal variables. It was hypothesized that associations of the bone characters should be greater with A but less with R compared with those for the tooth characters, and this was shown to be the case. The canonical correlation analysis also showed a strong association of P with G in each sex, although in a factor analysis these two variables associated with R.     

Growth and nutritional status of rural South African children 3–10 years old: The Ellisras growth study

This study presents cross-sectional data from an ongoing mixed-longitudinal study of growth of rural children from Ellisras, South Africa. The physical growth and nutritional status of 1,335 children (684 boys, 651 girls), 3–10 years of age, was determined using standard anthropometric techniques. Weight-for-age, height-for-age, and weight-for-height were expressed as Z-scores of the NHANES I and II or NCHS reference sample. A Z-score of less than −2 was used as the cut-off point to determine the prevalence of stunting and wasting. Mean heights increased parallel to the 50^th centile up to 6 years of age, thereafter both sexes diverged from the NHANES reference by approximately 0.5 cm per year. Mean weights followed a more consistent pattern from 3–7 years for both sexes, which was parallel to just below the 10^th centile, but diverged between 8 and 10 years of age. Z-scores of weight-for-height in both sexes varied between −1 to −2 throughout the age range and BMI values were lower than the 5^th centile of NHANES, indicating a significant amount of wasting within the sample. The sample exhibited a high prevalence of stunting, rising from less than 10% at 7 years to more than 30% by 10 years of age. Increments of the mean heights and weights indicate that the effects of stress may be a gradually accumulating process and that the growth increments of these children became increasingly poor in contrast to those of the reference sample. Since stunting in childhood is permanent, it may lead to a loss of physical work capacity in adulthood. Therefore, further investigation of the cause of poor growth among these rural children is imperative. Am. J. Hum. Biol. 12:42–49, 2000. © 2000 Wiley-Liss, Inc.     

Predictive equations for the estimation of body size in seals and sea lions (Carnivora: Pinnipedia)

Body size plays an important role in pinniped ecology and life history. However, body size data is often absent for historical, archaeological, and fossil specimens. To estimate the body size of pinnipeds (seals, sea lions, and walruses) for today and the past, we used 14 commonly preserved cranial measurements to develop sets of single variable and multivariate predictive equations for pinniped body mass and total length. Principal components analysis (PCA) was used to test whether separate family specific regressions were more appropriate than single predictive equations for Pinnipedia. The influence of phylogeny was tested with phylogenetic independent contrasts (PIC). The accuracy of these regressions was then assessed using a combination of coefficient of determination, percent prediction error, and standard error of estimation. Three different methods of multivariate analysis were examined: bidirectional stepwise model selection using Akaike information criteria; all‐subsets model selection using Bayesian information criteria (BIC); and partial least squares regression. The PCA showed clear discrimination between Otariidae (fur seals and sea lions) and Phocidae (earless seals) for the 14 measurements, indicating the need for family‐specific regression equations. The PIC analysis found that phylogeny had a minor influence on relationship between morphological variables and body size. The regressions for total length were more accurate than those for body mass, and equations specific to Otariidae were more accurate than those for Phocidae. Of the three multivariate methods, the all‐subsets approach required the fewest number of variables to estimate body size accurately. We then used the single variable predictive equations and the all‐subsets approach to estimate the body size of two recently extinct pinniped taxa, the Caribbean monk seal (Monachus tropicalis) and the Japanese sea lion (Zalophus japonicus). Body size estimates using single variable regressions generally under or over‐estimated body size; however, the all‐subset regression produced body size estimates that were close to historically recorded body length for these two species. This indicates that the all‐subset regression equations developed in this study can estimate body size accurately.     

TW3 bone age: RUS/CB and gender differences of percentiles for score and score increments

BACKGROUND: Longitudinal data on bone age progression is scarce. AIM: The study aimed to present reference values for Tanner-Whitehouse 3 (TW3) bone age score and score increments, and to provide means and standard deviations of appearance time for all TW3 stages. Gender differences and differences between radio ulna and short bones (RUS) and carpal bone (CB) scores were studied. SUBJECTS AND METHODS: Bone age data collected for ages 3 months to 20 years in 232 subjects during the First Zurich Longitudinal Study (1954-1976) were used. Smoothed empirical percentiles of TW3 RUS and CB scores for age, of score increments for age and of score increments for attained score are presented. Means and standard deviations of the appearance times are calculated by parametric censored regression. RESULTS: There are clear differences between the RUS and CB scores and between the genders. Boys are delayed with respect to girls, with different delays for RUS and CB. For RUS, differences in maturation reflect the known differences of physical growth, with a later and more intense peak in boys. For CB, there is little difference in timing and intensity. However, girls reach the final score about 2 years earlier than boys. The consistently earlier mean appearance times in girls indicate that skeletal maturation is, already in childhood, more rapid in girls than in boys. There are significant gender differences in the sequence of appearance. CONCLUSION: Reference values for TW3 score and score increments and mean appearance times for stages add to existing knowledge and indicate important RUS/CB and gender differences, whose sources are largely unknown.     

Orthodontic Study Cast Analysis—Reproducibility of Recordings and Agreement Between Conventional and 3D Virtual Measurements

Digital imaging of dental casts has become an alternative to conventional cast analysis in orthodontic treatment planning. The reproducibility of angular measurements made on virtual digital models has not as yet been evaluated. In order to study reproducibility for and agreement between a conventional analysis technique and virtual 3D imaging, 20 dental casts from subjects in the early mixed dentition, were measured twice with each technique by two orthodontists. Variables of interest were incisor rotation, angulation and irregularity, arch width, arch circumference, overjet, and overbite. Standard deviation (Dahlberg 1940) and coefficient of variation were used for evaluation of reproducibility. Mean differences between methods and examiners, correlation between mean differences and mean levels and 95% limits of agreement were used for describing systematic errors. The conventional technique showed less intraexaminer variation for angular variables than the 3D imaging method. Linear variables, with the exception of overbite, showed no clear trend as regards differences in reproducibility between the two methods. In general, examiner 1 showed less intraexaminer variation than examiner 2. The mean differences between the two methods expressed higher values for assessment of rotations with the O3DM method. Both angular and linear variables exhibited poor 95% limits of agreement. The conventional technique showed better overall reproducibility and thus appears to be more suitable for scientific work. However, the reproducibility attained by O3DM is clearly acceptable for clinical use. The two methods should not be used interchangeably.     

José Jordan

Background: A child's adult height is commonly predicted using their target height, based on mid-parent height. However, if no growth disorder is suspected, the child's current height is a far better predictor of their adult height.

Aim: To develop a chart to predict a child's adult height from their current height, adjusting for regression to the mean.

Subjects and methods: Data from the First Zurich Longitudinal Growth Study provided correlations between child height and adult height by age and sex, for use in a regression model predicting adult height centile from child height centile. The model was validated using data from the British 1946 and 1958 birth cohorts.

Results: The chart is illustrated superimposed on the British 1990 boys height chart. The predicted height has a standard error of 4–5 cm for ages from 4 years to puberty in both sexes. The regression adjustment partially compensates for biased predictions in early and late developers in puberty. A simplified version of the chart for restricted age ranges is also shown, as used on the UK-WHO 0–4 years growth charts.

Conclusion: The height prediction chart should be of value for parents, and indirectly professionals, to predict adult height in their children.     

Assessment of biological age by principal component analysis

A method of assessing biological age by the application of principal component analysis is reported. Healthy individuals (462) randomly selected from about 6000 men who had taken a 2-day health examination were studied. Out of the 30 physiological variables examined in routine check-ups, 11 variables were selected as suitable for the assessment of biological age based on the results of factor analysis and the physiological meaning of each test. This variable set was then submitted to principal component analysis, and the 1st principal component obtained from this analysis was used as an equation for assessing one's biological age. However, the biological age calculated from this equation is expressed as a score, so the estimated score was transformed to years (biological age) using the T-score idea. The biological age estimated by this method is practically useful and theoretically valid in contrast with the multiple regression model, because this approach eliminates and overcomes the following 2 big problems of the multiple regression model: (1) the distortion of the individual biological age at the regression edges; and (2) a theoretical contradiction in that a perfect model will merely be predicting the subject's chronological age, not his biological age.     

Pathway‐Guided Identification of Gene‐Gene Interactions

Assessing gene‐gene interactions (GxG) at the gene level can permit examination of epistasis at biologically functional units with amplified interaction signals from marker‐marker pairs. While current gene‐based GxG methods tend to be designed for two or a few genes, for complex traits, it is often common to have a list of many candidate genes to explore GxG. We propose a regression model with pathway‐guided regularization for detecting interactions among genes. Specifically, we use the principal components to summarize the SNP‐SNP interactions between a gene pair, and use an L1 penalty that incorporates adaptive weights based on biological guidance and trait supervision to identify important main and interaction effects. Our approach aims to combine biological guidance and data adaptiveness, and yields credible findings that may be likely to shed insights in order to formulate biological hypotheses for further molecular studies. The proposed approach can be used to explore the GxG with a list of many candidate genes and is applicable even when sample size is smaller than the number of predictors studied. We evaluate the utility of the proposed method using simulation and real data analysis. The results suggest improved performance over methods not utilizing pathway and trait guidance.     

Biochemical markers for detection of atherosclerotic vascular complications

Atherosclerotic vascular complication is considered as a final complication of lifestyle-related disease; however, suitable biochemical markers to detect vascular complications are yet to be determined. To clarify the clinical usefulness of oxidized low-density lipoprotein(oxLDL), high-sensitive C-reactive protein (hsCRP), and serum lysophosphatidylcholine (LPC) for the detection of vascular complications in type 2 diabetics, we evaluated the clinical implications of oxLDL, hsCRP and LPC in relation to clinical backgrounds and vascular complications. OxLDL was measured by ELISA(Mercodia, Sweden), hsCRP by immunonephelometry (Roche Diagnostics, Germany) and LPC by enzymatic methods (Alfresa, Japan), respectively. The oxLDL level did not differentiate any vascular complications; however, hsCRP was significantly higher in patients with ischemic heart disease (IHD), and LPC was significantly lower in patients with IHD and macroangiopathy composed of IHD, cerebral vascular accident and arteriosclerosis obliterans. Interestingly, the plasma LPC level was lower in women than in men. Multivariate regression analysis revealed that IHD potently contributed negatively to the LPC level, and also contributed positively to the hsCRP level, but did not contribute to the oxLDL level. Multivariate regression analysis also revealed that LPC, but not oxLDL nor hsCRP, contributed to the development of IHD and macroangiopathy. Thus, LPC is a better biochemical marker than oxLDL and hsCRP for the detection of vascular complications.     

Principal component analysis of shape variables in adult individuals

Human body shape variables were obtained by adjusting 34 distances between trunk/limbs, and head/face, landmarks, for an overall appropriate body size measurement. The adjustment was based on regression analysis. Principal component analysis was applied to thus defined shape variables to obtain shape dimensions in 99 normal adult males and 103 females. The first principal component either for trunk/limbs shape variables, or for head/face variables (considered separately in the analyses) is similar in both sexes in that it represents relative proportions between trunk and limb lengths and widths, and between midfacial lengths and widths, respectively. However there are appreciable differences in the succeeding components. The problem of interpretation of body shape dimensions, especially those accounting for less than 20% of the sample variance, as well as difficulties in assessing the biological meaning of dependence structures determined by principal component analysis in humans, are discussed.