A two-part model for reference curve estimation subject to a limit of detection.

Reference curves are commonly used to identify individuals with extreme values of clinically relevant variables or stages of progression which depend naturally on age or maturation. Estimation of reference curves can be complicated by a technical limit of detection (LOD) that censors the measurement from the left, as is the case in our study of reproductive hormone levels in boys around the time of the onset of puberty. We discuss issues with common approaches to the LOD problem in the context of our pubertal hormone study, and propose a two-part model that addresses these issues. One part of the proposed model specifies the probability of a measurement exceeding the LOD as a function of age. The other part of the model specifies the conditional distribution of a measurement given that it exceeds the LOD, again as a function of age. Information from the two parts can be combined to estimate the identifiable portion (i.e. above the LOD) of a reference curve and to calculate the relative standing of a given measurement above the LOD. Unlike some common approaches to LOD problems, the two-part model is free of untestable assumptions involving unobservable quantities, flexible for modeling the observable data, and easy to implement with existing software. The method is illustrated with hormone data from the Third National Health and Nutrition Examination Survey.

Association of inflammation with worsening HOMA-insulin resistance.

AIMS/HYPOTHESIS: We examined the cross-sectional and longitudinal relationships between C-reactive protein (CRP), a marker of low-grade inflammation, and insulin resistance and whether the association was independent of obesity and oxidative stress. METHODS: CRP and insulin resistance (homeostasis model assessment of insulin resistance [HOMA-IR]) data were obtained in a population-based, prospective observational study, Coronary Artery Risk Development in Young Adults (CARDIA), during 1992-2006. RESULTS: CRP showed a significant positive association with insulin resistance, both cross-sectionally and longitudinally (5 year follow-up). The estimated increment in HOMA-IR was 0.34 log(e)(pmol/l x [mmol/l]/156.25) (p value for trend <0.0001) in the highest vs lowest CRP quartiles in cross-sectional analysis, whereas the corresponding estimate was 0.12 (p trend <0.0001) in the highest vs lowest CRP quartiles longitudinally over 5 years. The gradient of HOMA-IR across CRP was attenuated but remained statistically significant after controlling for body fat measurements (0.06 in the highest vs lowest CRP in both cross-sectional [p value for trend = 0.001] and longitudinal analyses [p value for trend = 0.01]), and was little changed by further adjustment for oxidative stress markers (F(2)-isoprostanes and oxidised LDL). There were consistent increments in the levels of HOMA-IR with increasing concentrations of CRP over time. In contrast, higher HOMA-IR did not predict future increases in CRP. Findings were similar using fibrinogen as the predictor variable. CONCLUSIONS/INTERPRETATION: Although a substantial portion of this association was explained by obesity, CRP was independently related to concurrent and future insulin resistance.

Growth and endocrine function in patients with Hurler syndrome after hematopoietic stem cell transplantation.

Short stature is characteristic of Hurler syndrome, or mucopolysaccharidosis type IH (MPS IH). Hematopoietic stem cell transplantation (HSCT) is used to treat children with MPS IH. While HSCT corrects some of the metabolic features of MPS IH, its effects on growth are not well delineated. We investigated growth in patients with MPS IH after HSCT and described accompanying endocrine abnormalities. A cohort of 48 patients with MPS IH who had received HSCT between 1983 and 2005 were included. The prevalence of short stature (height <-2 s.d. score, SDS) before HSCT was 9%, and increased to 71% at last follow-up (6.9+/-5.1 years after HSCT). Short stature was positively associated with increased age at HSCT (P=0.002) and TBI (P=0.009). In total, 23% had growth hormone deficiency and/or low insulin-like growth factor-1, one female patient had premature adrenarche, one precocious puberty and 27% had clinical or subclinical hypothyroidism. Growth failure is highly prevalent in children with MPS IH after HSCT. Children who had no TBI exposure and were younger at the time of HSCT had a better height outcome.

Incremental growth charts.

Serial data from 818 normal American white children have been used to construct 6-month increment charts for weight (0 to 18 years), head circumference and recumbent length (0 to 3 years), and stature (2 to 18 years). The attained growth status of these children is very close to that shown in the National Center for Health Statistics charts at all ages. Consequently, the present increment charts are widely applicable; it is expected secular changes will not affect their usefulness in ther near future. The need for accurate anthropometric techniques is emphasized and guides to the interpretation of findings are presented.

Guidelines for overweight in adolescent preventive services: recommendations from an expert committee. The Expert Committee on Clinical Guidelines for Overweight in Adolescent Preventive Services.

An expert committee was convened to determine specific criteria for overweight to be integrated into routine preventive screening of adolescents. Body mass index (BMI) should be used routinely to screen for overweight adolescents. Youth with BMIs > or = 95th percentile for age and sex, or > 30 (in kg/m2) should be considered overweight and referred for indepth medical follow-up to determine underlying diagnoses. Adolescents with BMIs > or = 85th percentile but < 95th percentile or < or = 30, should be considered at risk of overweight, and should be referred to a second-level screen. The second-level screen includes family history, blood pressure, total cholesterol, large prior increment in BMI, and concern about weight. If youths are positive for any of the items on the second-level screen they should be referred for further medical assessment.

Fat areas as estimates of total body fat.

The efficacy of cross-sectional fat areas in estimating total body fat was investigated in a sample of white American children and adults. Body density and total fat weight (kg) in the body were determined by hydrostatic weighing. Fat areas were calculated for the arm and calf using the appropriate limb circumferences and skinfolds measured at the triceps, biceps, and calf sites; also, a fat area was calculated using the average of triceps and biceps sites and arm circumference. Cross-sectional fat areas do not estimate body density (and percentage fat) any better than the corresponding skinfolds. In estimating weight of fat in the body, however, fat areas are systematically better estimators than corresponding skinfold thicknesses.

Incremental growth tables: supplementary to previously published charts.

Serial data from 818 normal white American children from the Fels Longitudinal Study have been used to construct tables of 6-mo increments for weight (0 to 18 yr), stature (3.0 to 18 yr), head circumference, and recumbent length (0 to 3 yr). The tables supplement previously published incremental growth charts. The attained growth status of these children has been shown to be very close to that displayed in the National Center for Health Statistics tables at all ages, and secular trends in the data have been shown to be minimal. Consequently, the present tables should be widely applicable for some time in the future. Guidelines for the use of the tables and interpretation of findings are presented.

Compressibility of skinfolds and the measurement of subcutaneous fatness.

Compressibility of subcutaneous fat thickness when measured with skinfold calipers was investigated in 65 white American youths. Compression of skinfolds was determined relative to measurements of subcutaneous fat thickness from radiographs at each of seven sites. There is statistically significant heterogeneity among sites in skinfold compression, with skinfolds on the medial and lateral calf being the least compressible of those measured. There is little statistically significant sex difference in skinfold compression in the present sample and, within the range concerned, there were no significant correlations between skinfold compressibility and age. When compressibilities of the seven skinfolds were intercorrelated within individuals, statistically significant average correlations were obtained, indicating that individuals tend toward similar degrees of skinfold compressibility among sites. This communality of skinfold compression within individuals is such that, at least in male youth, there are significant differences among individuals in the average compressibility of the seven skinfolds.

Nutritional supplementation, psychosocial stimulation, and growth of stunted children: the Jamaican study.

The benefits of nutritional supplementation, with or without psychosocial stimulation, on the growth of stunted children were evaluated. Children aged 9-24 mo with lengths less than -2 SD of the National Center for Health Statistics references (n = 129) were randomly assigned to four groups: control, nutritional supplementation, stimulation, and both interventions. A fifth group with lengths greater than -1 SD was also enrolled. Length, weight, head and arm circumferences, and triceps and subscapular skinfold thicknesses were measured on enrollment and 6 and 12 mo later. Multiple-regression analysis was used to determine the effects of the interventions in which age, sex, initial status, initial dietary intake, and several socioeconomic variables were controlled for. Stimulation had no effect on growth and there was no interaction between the interventions. After 12 mo supplemented children had significantly increased length, weight, and head circumference (all P less than 0.01). The effects of supplementation were not cumulative but occurred in the first 6 mo.

Morbidity and the growth of stunted and nonstunted children, and the effect of supplementation.

Children aged 9-24 mo were recruited by a survey of poor areas of Kingston, Jamaica. Stunted children were randomly assigned to supplementation or not. Weekly morbidity histories were taken for 2 y. Separate multiple regressions on each symptom for weight or length gain in 2-mo intervals showed significant reductions in weight gain with coughing, apathy, anorexia, diarrhea, and fever, ranging from -2.1 to -16.8 g/d ill. Apathy and diarrhea reduced gains in length (-0.26 and -0.20 mm/d ill). Significant reductions in linear growth with lower respiratory-tract infections (-0.16 mm/d ill) occurred only in nonsupplemented children. Growth over 4-mo intervals was reduced if diarrhea occurred in the first 2 mo of the interval but there were no long-term effects of apathy, fever, or anorexia. Some of the effects of morbidity on growth were therefore transient and morbidity is unlikely to be a major cause of growth retardation in this population.

Body composition assessment in American Indian children.

Although the high prevalence of obesity in American Indian children was documented in several surveys that used body mass index (BMI, in kg/m2) as the measure, there is limited information on more direct measurements of body adiposity in this population. The present study evaluated body composition in 81 boys (aged 11.2+/-0.6 y) and 75 girls (aged 11.0+/-0.4 y) attending public schools in 6 American Indian communities: White Mountain Apache, Pima, and Tohono O'Odham in Arizona; Oglala Lakota and Sicangu Lakota in South Dakota; and Navajo in New Mexico and Arizona. These communities were participating in the feasibility phase of Pathways, a multicenter intervention for the primary prevention of obesity. Body composition was estimated by using a combination of skinfold thickness and bioelectrical impedance measurements, with a prediction equation validated previously in this same population. The mean BMI was 20.4+/-4.2 for boys and 21.1+/-5.0 for girls. The sum of the triceps plus subscapular skinfold thicknesses averaged 28.6+/-7.0 mm in boys and 34.0+/-8.0 mm in girls. Mean percentage body fat was 35.6+/-6.9 in boys and 38.8+/-8.5 in girls. The results from this study confirmed the high prevalence of excess body fatness in school-age American Indian children and permitted the development of procedures, training, and quality control for measurement of the main outcome variable in the full-scale Pathways study.

Subcutaneous fat distribution in males and females from 1 to 39 years of age.

Subcutaneous fat thicknesses measured in two longitudinal studies, the Denver Growth Study and the Melbourne Growth Study, were examined to determine 1) if one site of subcutaneous fat measurement is more representative than another of the body's subcutaneous fat layer, 2) if some measurements of subcutaneous fat from specific body areas are more representative of the subcutaneous fat layer than others, and 3) if there are sex and age differences in intersite relationships. Data from the Denver Study include skinfold thicknesses from 10 sites measured serially at annual ages from 4 to 39 yr and radiographic fat thickness measured at five sites at annual ages from 1 to 18 yr. In the Melbourne Study, five skinfold thicknesses were measured serially in children from 2 to 15 yr of age. Relatively low intersite communalities during the prepubertal years suggest a tendency in each sex for considerable site-to-site variation during this period. However, immediately before puberty and throughout adolescence, high communalities indicate that the thickness of subcutaneous fat at any site is highly related to thickness at all other sites. After puberty and into the mid-20's, there is a reoccurrence of greater site-to-site variability. This higher degree of variation continues into early middle age only in women, implying that more changes occur in their subcutaneous fat, that differentially affect various parts of the body, than in men.

A method to estimate prevalence of iron deficiency and iron deficiency anemia in adolescent Jamaican girls.

A method is presented to estimate a cutoff for hemoglobin concentration appropriate for estimating the prevalence of iron deficiency anemia in poor Jamaican girls 13-14 y of age. Iron deficiency was determined from a three-variable model of iron status (serum ferritin, erythrocyte protoporphyrin, and mean corpuscular volume). The most appropriate hemoglobin cutoff was considered the one that minimized misclassification of iron deficiency: that yielding the maximum kappa coefficient for correctly classifying iron deficiency between 100 and 120 g/L, at 1-g/L intervals. By using this method, a hemoglobin cutoff of 107 g/L was considered most appropriate. This cutoff and the other indicators were used to estimate prevalence of iron deficiency and iron deficiency anemia in the Jamaican girls: 7.6% and 4.3%, respectively. This approach should be appropriate for determining hemoglobin cutoffs for iron deficiency anemia in other populations.

Prevalence of obesity in American Indians and Alaska Natives.

Obesity is an important risk factor for cardiovascular diseases and non-insulin-dependent diabetes, which are chronic diseases that afflict American Indians and Alaska Natives today. Because American Indians are not represented in most national health and nutrition surveys, there is a paucity of data on actual prevalence of obesity in American Indians. We estimated prevalence of overweight and obesity for American Indian adults, school-age children, and preschool children from existing data. The prevalence of obesity in adults was estimated from self-reported weights and heights obtained from a special survey of American Indians performed as part of the 1987 National Medical Expenditure Survey. Prevalence of obesity in American Indians was 13.7% for men and 16.5% for women, which was higher than the US rates of 9.1% and 8.2%, respectively. Obesity rates in American Indian adolescents and preschool children were higher than the respective rates for US all-races combined.

Design and statistical analysis for the Pathways study.

We report the design, rationale, and statistical procedures used in Pathways, a randomized, school-based intervention for the primary prevention of obesity in American Indian children. The intervention, which is now being implemented in 7 American Indian communities around the country, includes a health-promotion curriculum, a physical education program, a school meal program, and a family involvement component. Forty-one schools serving American Indian children were randomly assigned to be either intervention or control groups. The intervention will begin in the third grade and continue through the end of the fifth grade. Efficacy of intervention will be assessed by differences in mean percentage body fat, calculated by a prediction equation, between intervention and control schools at the end of the fifth grade. Power computations indicate that the study has power to detect a mean difference of 2.8% in body fat. Data analysis will use intention-to-treat concepts and the mixed linear model. The study will be completed in 2000.

Parent-specific adjustments for evaluation of recumbent length and stature of children.

A clinical "parent-specific" method for evaluating recumbent length and stature of children, allowing for the stature of parents, is presented. This method uses midparent stature and allows adjustment of recumbent length (birth to 3 years) and stature (3 to 18 years) of US children relative to National Center for Health Statistics growth charts. The method is based on parent-child relationships for 586 midparent-child pairs participating in the Fels Longitudinal Study, and on more than 16,000 serial measurements of recumbent length and stature. The method provides a clinical tool to separate the normal genetic contribution of parental stature to stature of children from other factors that affect stature such as malnutrition or disease.

Differences and trends in antioxidant dietary intake in smokers and non-smokers, 1980-1992: the Minnesota Heart Survey.

PURPOSE: Differences and secular trends in dietary antioxidant vitamin intake (vitamins E, C, and beta-carotene) in current non-smokers, light smokers, and heavy smokers were examined as part of the Minnesota Heart Survey. METHODS: Three cross-sectional surveys were conducted in adults ages 25-74 years in 1980-82 (N = 1682), 1985-87 (N = 2326), and 1990-92 (N = 2487). Dietary information was obtained from a 24-hour dietary recall. Smoking was assessed through self-report. Intakes were adjusted for age, energy intake, body mass index, education level, and exercise level (vitamins E, C and beta-carotene). RESULTS: Antioxidant vitamin intakes were significantly higher in non-smokers than in light (1-20 cig/day) and heavy smokers (>20 cig/day) when all three survey periods were combined. In men, mean vitamin E intake was 9.2 mg, 8.6 mg, and 7.8 mg for non-smokers, light smokers, and heavy smokers, respectively. Results were similar in men for beta-carotene (non-smokers 1408 microg, light smokers 1287 microg, and heavy smokers 1064 microg), and vitamin C (non-smokers 81 mg, light smokers 67 mg, and heavy smokers 56 mg). Women had results of similar magnitude and direction. From 1980-92, secular trends in men showed non-significant increases from 1980-82 to 1990-92 in beta-carotene (+6.1%), while decreases were observed in vitamins E (-1.1%) and C (-2.6%). In contrast, women had large decreases in all antioxidant vitamin intakes: vitamin E (-13%), vitamin C (-18.6%), and beta-carotene (-16.2%). CONCLUSIONS: Light and heavy smokers had a significantly lower overall mean dietary antioxidant vitamin intake than non-smokers. Over the decade, antioxidant dietary intake remained relatively stable in men and decreased in women in Minneapolis-St. Paul, despite improvements in access to antioxidant rich fruits and vegetables.

Age-associated statural loss and socioeconomic status.

Association between socioeconomic score (SES) and age-related statural loss were investigated in a cross-sectional sample of 718 adults from rural Colombia. Estimates of statural aging were calculated by estimating the maximum statural attainment of each subject from statural relationships with subischial length. Statural loss was correlated with SES, r=-0.06 (not significant) for men, and r=-0.14 (p less than 0.01) for women. Subjects with a high SES tended to lose stature at a slower rate than those with a low SES. For women, the difference in rates of statural loss between the low SES groups was significant (p less than 0.05).

A United States national reference for fetal growth.

OBJECTIVE: To develop a current national fetal growth curve that can be used as a common reference point by researchers to facilitate investigations of the predictors and consequences of small and large for gestational age delivery. METHODS: Single live births to United States resident mothers in 1991 (n = 3,134,879) were used for the development of this curve, which was compared with four previously published fetal growth curves. Techniques were developed to address cases with implausible birth weight-gestational age combinations and to smooth fetal growth curves across gestational age categories. RESULTS: In general, the previously published fetal growth curves underestimated the 1991 United States reference curve. This underestimation is most apparent during the latter weeks of gestation, approximately 33-38 weeks. CONCLUSION: Our findings indicate that the prevalence of fetal growth restriction (FGR) will vary markedly, depending on the fetal growth curve used. Furthermore, many previously published fetal growth curves no longer provide an up-to-date reference for describing the distribution of birth weight by gestational age and for determining FGR that is consistent with the most recent live birth data for the entire United States.

Pregnancy complications and birth outcomes in obese and normal-weight women: effects of gestational weight change.

OBJECTIVE: To compare the pregnancy course and outcomes in obese and normal-weight women and their associations with gestational weight change. METHODS: Multivariate logistic regression described the relation of weight change to pregnancy course and outcomes in a retrospective study of 683 obese and 660 normal-weight women who delivered singleton living neonates. RESULTS: Compared with normal-weight women, obese women gained an average of 5 kg (11 lb) less during pregnancy and were more likely to lose or gain no weight (11% versus less than 1%). Obese women were significantly more likely to have pregnancy complications, but the incidence of complications was not associated with weight change. Compared with obese women who gained 7-11.5 kg (15-25 lb), obese women who lost or gained no weight were at higher risk for delivery of infants under 3000 g or small for gestational age infants, and those who gained more than 16 kg (35 lb) were at twice the risk for delivery of infants who were 4000 g or heavier. CONCLUSION: Gestational weight change was not associated with pregnancy complications in obese or normal-weight women. To optimize fetal growth, weight gains of 7-11.5 kg (15-25 lb) for obese women and 11.5-16 kg (25-35 lb) for normal-weight women appear to be appropriate.