The association of variants in the FTO gene with longitudinal body mass index profiles in non-Hispanic white children and adolescents.

OBJECTIVE: To investigate possible age-related changes in associations between polymorphisms in the fat mass and obesity-associated (FTO) gene and higher body mass index (BMI). DESIGN AND SUBJECTS: Multilevel mixed regression models were used to examine associations between four FTO variants and longitudinal BMI profiles in non-Hispanic white and African American children and adolescents 8-17 years of age from two different longitudinal cohort studies, the Bogalusa Heart Study (BHS) and Project HeartBeat! (PHB). In the BHS, there were 1551 examinations of 478 African Americans and 3210 examinations of 1081 non-Hispanic whites; in PHB, there were 971 examinations of 131 African Americans and 4458 examinations of 505 non-Hispanic whites. RESULTS: In African Americans, no significant FTO associations with BMI were found. In non-Hispanic whites, linkage disequilibrium among all four variants made haplotype analysis superfluous, so we focused on the single-nucleotide polymorphism, rs9939609. In longitudinal multilevel models, the A/A genotype of rs9939609 was associated with higher BMI in non-Hispanic whites in both cohorts at all ages. A significant age-by-genotype interaction found only in the BHS cohort predicted that in those with the A/A genotype, BMI would be ∼0.7 kg m(-2) higher at age 8 and ∼1.6 kg m(-2) higher at age 17 than in those with A/T or T/T genotypes. The design of PHB limited follow-up of any single individual to 4 years, and may have reduced the ability to detect any age-by-genotype interaction in this cohort. CONCLUSIONS: The A/A genotype of rs9939609 in the FTO gene is associated with higher longitudinal BMI profiles in non-Hispanic whites from two different cohorts. The association may change with age, with the A/A genotype being associated with a larger BMI difference in late adolescence than in childhood, though this was observed only in the BHS cohort and requires verification.

Nutrient intake of obese female binge eaters.

OBJECTIVE: To compare the 6-month change in selected nutrients and number of binge days (from 7-day food records) between obese binge eaters randomly assigned to either a behavioral self-management (BSM) or waiting list control (WLC) group. Within each of the 2 groups, the average intake of selected nutrients on binge and nonbinge days at baseline and at 6 months were compared. DESIGN: A randomized, controlled, intervention study with assessments at entry and 6 months later. SUBJECTS: Forty-six women in the BSM group and 36 in the WLC group completed the 6-month measurement. Participants were 25 to 50 years of age, 30 to 90 pounds overweight, did not have a history of physical or psychological illnesses, and scored 20 or greater on the binge eating scale. INTERVENTION: Participants in the BSM intervention received 6 months of weekly, 1-hour classes taught by registered dietitians. Participants in the WLC group were not contacted during the 6 months. OUTCOME MEASURES: The main outcome measures were change in energy consumed (kilocalories); percentage of energy from fat, protein, and carbohydrate; grams of fiber/1,000 kcal; and change in the number of self-reported binge days. STATISTICAL ANALYSES: Weight at 6 months was compared using a 2-sample t test. The change in the number of binge days at 6 months and the amount of change in selected nutrients by group was compared using the 2-sample t test. The paired t test was used to compare the average nutrient intakes on binge and nonbinge days within groups. RESULTS: No significant difference was found in the 6-month change between groups in any of the selected nutrients. The BSM group reported a greater reduction in binge days between baseline and 6 months compared with the WLC group (mean 1.0 vs 1.7, P < 0.03). Within the BSM group at 6 months, energy intake and percentage of energy from fat on nonbinge days were significantly reduced compared with binge days. At baseline within the WLC group, energy intake increased and percentage of energy from protein decreased significantly on nonbinge days compared with binge days. Within the WLC group at 6 months, energy intake and percentage of energy from fat significantly decreased and percentage of energy from protein significantly increased on nonbinge days. CONCLUSIONS: Our results suggest that collecting dietary information from participants identified with binge eating disorder is challenging. Dietitians who conduct behavioral weight management programs may require additional training in identifying and understanding the psychological characteristics of participants with binge-eating disorder.

Multilevel Models and Unbiased Tests for Group Based Interventions: Examples from the Safer Choices Study.

For many large-scale behavioral interventions, random assignment to intervention condition occurs at the group level. Data analytic models that ignore potential non-independence of observations provide inefficient parameter estimates and often produce biased test statistics. For studies in which individuals are randomized by groups to treatment condition, multilevel models (MLMs) provide a flexible approach to statistically evaluating program effects. This article presents an explanation of the need for MLM's for such nested designs and uses data from the Safer Choices study to illustrate the application of MLMs for both continuous and dichotomous outcomes. When designing studies, researchers who are considering group-randomized interventions should also consider the features of the multilevel analytic models they might employ.

Factors influencing success of monovision contact lens fitting: survey of contact lens diplomates.

BACKGROUND: Contact lenses are frequently fitted so that one eye is focused for distance and the other for near (monovision) in order to reduce the dependence of presbyopic patients on bifocal spectacle correction. METHODS: We surveyed Diplomates in the Cornea and Contact Lens section of the American Academy of Optometry (N = 179) regarding their estimates of success with monovision fitting and factors influencing their perception of success. Results were statistically analyzed to determine the fitting philosophies of responders estimating that they achieve high success rates (HSRs) and low success rates (LSRs), respectively. RESULTS: Completed surveys were returned by 98 Diplomates (54.8%), who estimated an overall success rate of approximately 71.7% with monovision; 50.5% "totally successful" (wear monovision correction full-time without problems or need for additional correction) and 21.2% "successful" (wear monovision full-time, but experience some symptoms of blur or other discomfort, may wear over-correction at times for driving or detailed near work). Sighting dominance was the most frequently used method to determine the distant eye, followed by the plus-lens test. The HSR group tended to be more likely to take occupational factors into account when deciding which eye to use for the distance lens, and they used more flexibility in determining the near-add power. The LSR group discontinued monovision sooner than the HSR group, who continued treatment for longer than 4 weeks during the adaptation period much more often for all add ranges. CONCLUSIONS: Practitioners who regard monovision treatment as very successful have somewhat different fitting philosophies than those who regard monovision as unsuccessful. Implementation of highly successful practitioners' fitting techniques may increase success with monovision contact lens fitting.

Gender and treatment differences in knowledge, health beliefs, and metabolic control in Mexican Americans with type 2 diabetes.

PURPOSE: The purpose of this project was to describe metabolic control, knowledge, and health beliefs of Mexican Americans with type 2 diabetes. METHODS: The study site was Starr County, Texas, a border community located on the Rio Grande River and bordering northern Mexico. Of the total sample of 360 persons, 252 agreed to participate in this intervention study and were randomized either to the treatment group or the control group that waited 1 year to begin the intervention. RESULTS: The majority of individuals were Spanish-speaking females with a mean age of 54 years and a mean diabetes duration of 8 years. For those treated with diet only, males exhibited higher fasting blood glucose levels than females. Gender effects were seen for cholesterol level, with females exhibiting higher levels than males. Males expressed stronger perceptions of control and social support for diet. Bivariate relationships were found between acculturation and diabetes knowledge. The health belief subscales of control and impact on job together explained 16% of the variance in HbA1c values. CONCLUSIONS: Males and females held differing beliefs about ability to control their diabetes and degree of social support for diet. The impact of gender differences on ability to integrate diabetes self-care and on effectiveness of diabetes programs has not been determined but should be considered in future research.

Validity of echocardiographic measurement in an epidemiological study. Project HeartBeat!

In Project HeartBeat!, a longitudinal study of cardiovascular disease risk factors in healthy children and adolescents, 3 samples of 40, 80, and 182 echocardiograms, respectively, were randomly selected and reread to evaluate intraobserver and interobserver variabilities and comparability between measurements of field echocardiographic technicians and reference readings at Texas Children's Hospital. Included in the evaluation were 8 M-mode echocardiographic measurements, ie, aortic root diameter, left atrial diameter, and end-diastolic and end-systolic measurements of interventricular septal thickness, left ventricular (LV) diameter, and LV posterior wall thickness; 8 Doppler measurements; and a calculated LV mass. Means and SDs of the differences of the paired measurements were used to assess the relative bias and random error of the measurements. For the intraobserver comparison, means and SDs of the differences were very small, indicating that the echo measurements were performed consistently by each project echo technician. Interobserver comparison showed statistically but not clinically significant differences between the paired readings of end-diastolic septal thickness, end-systolic LV posterior wall thickness, and 5 Doppler measurements. Comparison with reference readings at Texas Children's Hospital showed significant differences in diastolic LV diameter, systolic septal thickness, and right ventricular ejection time. These differences, however, were minimal with limited clinical significance. Mean differences in LV mass for the corresponding comparisons were -1.82, 4.50, and 0.0013 g, and the SDs were 18.79, 24.16, and 12.35 g, respectively. We conclude that the echocardiographic measurements taken from healthy children in a longitudinal study can be made accurately with acceptable reproducibility.

Development of cardiovascular risk factors from ages 8 to 18 in Project HeartBeat! Study design and patterns of change in plasma total cholesterol concentration.

BACKGROUND: Project HeartBeat! is a longitudinal study of the development of cardiovascular risk factors as growth processes. Patterns of serial change, or trajectories, from ages 8 to 18 years for plasma total cholesterol concentration (TC) and percent body fat illustrate the design and synthetic cohort approach of the study. METHODS AND RESULTS: Six hundred seventy-eight children (49.1% female, 20.1% black) entered the study at ages 8, 11, and 14 years and were followed up with examinations every 4 months for < or = 4 years. Multilevel analysis demonstrated trajectories for population mean values of TC and percent body fat in sex-specific synthetic cohorts from ages 8 to 18 years. Polyphasic patterns of change in TC were confirmed, with notable sex differences in age patterns and with minimum mean values of TC of 3.85 mmol/L for females and 3.59 for males. As illustrated by data for males, the approximate 75th percentile values of mean TC ranged from 4.78 mmol/L at its early peak to 4.06 at its late-teen nadir. Percent body fat exhibited a trajectory closely parallel with that for TC only for males and appeared to be unrelated for females. CONCLUSIONS: The polyphasic trajectory for TC from ages 8 to 18 years differs between females and males, indicates marked age variation in 75th percentile values and, in males only, closely parallels the trajectory for percent body fat. These and other results indicate the value of both follow-up every 4 months across age intervals to detect rapid risk factor change and the synthetic cohort approach for gaining new insights into the dynamics and possible determinants of this change from ages 8 to 18 years.

Low birth weight and blood pressure at age 7-11 years in a biracial sample.

The hypothesis that birth weight predicts blood pressure inversely at age 7 through 11 years was examined in 1,446 white children and black children in Washington Parish, Louisiana. Two data sets of the Bogalusa Heart Study were merged: 1) newborn cohort participants (n = 233), initially examined at birth, 1973-1974, and reexamined in 1984-1985 at ages 9 through 11 years; and 2) subjects examined at ages 7 through 11 years in 1987-1988 (n = 1,213) whose birth weight was collected from birth certificates in 1991. The prevalence ratios for being in the race-, sex-, and age-specific upper decile of diastolic blood pressure in children born with low birth weight (< 2,500 g) versus those with birth weight > or = 2,500 g were 0.85 (95% confidence interval 0.28-2.56) for white boys, 2.66 (95% confidence interval 1.24-5.70, p < 0.05) for black boys, 1.38 (95% confidence interval 0.63-3.03) for white girls, and 1.05 (95% confidence interval 0.40-2.75) for black girls. For systolic blood pressure, the corresponding prevalence ratio for each race-sex group did not differ from one. When the analyses were restricted to full-term births, prevalence ratios in any race-sex group did not differ from one for systolic and diastolic blood pressure. In multiple linear regression analyses, the concurrently determined Quetelet index (p < 0.001) was a much stronger correlate of systolic and diastolic blood pressure after appropriate adjustment than was birth weight (p > 0.05). From this study, there is some evidence that low birth weight may determine a risk for subsequent high blood pressure in black boys in the age group 7 through 11 years, but the inconsistency of the results for other race-sex groups was unexpected and remains unexplained, if the underlying hypothesis is true.

Low birth weight and serum lipid concentrations at age 7-11 years in a biracial sample.

The relation between birth weight and serum lipid concentrations at age 7 through 11 years was examined in a sample of 1,411 black children and white children in Washington Parish, Louisiana. Two data sets of the Bogalusa Heart Study were merged: 1) newborn cohort participants (n = 225) initially examined at birth, 1973-1974, and reexamined in 1984-1985 at age 9 through 11 years; and 2) subjects examined at ages 7 through 11 years in 1987-1988 (n = 1,186) whose birth weight was collected from birth certificates in 1991. The prevalence ratios for being in the race-, sex-, and age-specific upper decile of serum lipid concentrations in children born with low birth weight (< 2,500 g) versus those with birth weight > or = 2,500 g were calculated per race-sex group. Among white boys with low birth weight, higher than expected percentages of subjects were in the highest decile group of triglyceride concentrations (0.01 < p < 0.05). The prevalence ratio was 2.42 (95% confidence interval 1.19-4.91). When premature infants were excluded, only for white girls was a greater than expected percentage of subjects with low birth weight found to be in the highest decile group of triglyceride concentrations. The corresponding prevalence ratio for white girls was 3.23 (95% confidence interval 1.16-9.00). In analyses that either included or excluded premature infants, prevalence infants, prevalence ratios for triglyceride concentrations in black boys and black girls and for the low density lipoprotein cholesterol/high density lipoprotein cholesterol ratio, total cholesterol concentration, and very low density lipoprotein cholesterol concentration in all race-sex groups were not significantly different from one. To our knowledge, this is the first study finding associations between low birth weight and elevated triglyceride concentrations in later childhood. A follow-up study among adults is recommended.

Diffusion of an effective tobacco prevention program. Part II: Evaluation of the adoption phase.

This paper presents the results of theory-based intervention strategies to increase the adoption of a tobacco prevention program. The adoption intervention followed a series of dissemination intervention strategies targeted at 128 school districts in Texas. Informed by Social Cognitive Theory, the intervention provided opportunities for districts to learn about and model themselves after 'successful' school districts that had adopted the program, and to see the potential for social reinforcement through the knowledge that the program had the potential to have an important influence on students' lives. The proportion of districts in the Intervention condition that adopted the program was significantly greater than in the Comparison condition (P < 0.001). Stepwise logistic regression indicated that the variables most closely related to adoption among intervention districts were teacher attitudes toward the innovation and organizational considerations of administrators. Recommendations for the development of effective strategies for the diffusion of innovations are presented.

Factors associated with serum cholesterol level in a pediatric practice. Cholesterol screening in a pediatric practice.

The associations between age, sex, height, Quetelet index, blood pressure, and serum cholesterol level were examined among 1406 routinely screened children, aged 4 to 19 years, in a pediatric practice. After adjustment for sex and age, height and Quetelet index were associated with serum cholesterol levels. Quetelet index was shown by multiple linear regression to be positively related to cholesterol levels (b = 0.780, P < 0.01), but the predictive value of screening based on an elevated Quetelet index was marginal. Clustering of elevated serum cholesterol level, Quetelet index, and systolic blood pressure was observed. Familial aggregation of cholesterol levels was demonstrated using analysis of variance for 742 children from 342 families included in the regression analysis (F341,400 = 1.56, P < 0.0001). The intraclass correlation coefficient, a measure of familial aggregation, was 0.205 (P < 0.0001). Age, sex, height, Quetelet index, and familial aggregation accounted for 10.6% of the variance in serum cholesterol levels. Siblings of children with high cholesterol levels are a high-yield group in cholesterol screening.

Is body fat distribution associated with cardiovascular risk factors in childhood?

The authors studied the association of cardiovascular risk variables with body fat distribution (BFD) in a cross-sectional sample of 743 Texas schoolchildren of both sexes ages 6-14 years. This association is well known in adults and several useful indices of BFD are available. Whether they are applicable to children remains a question of importance for epidemiological investigations in this age group. Canonical correlations between anthropometric (five skinfolds, four circumferences) and risk (blood pressures, cholesterol, pulse) variables ranged from 0.37 to 0.82 depending on sex/age group (p < 0.01). The skinfold vector suggested an association of risk with central fat at most but not all ages. The circumference vectors, on the other hand, suggest that size or fatness, not BFD, was related to risk. Partial correlation and stepwise regression of fatness and BFD indices with cardiovascular risk factors as dependent variables, showed that 'fatness' or 'size' was more clearly associated with risk factors than BFD. The variables most consistently entering the regression were hip circumference and sum of skinfolds. These results contrast with studies of adults or sexually mature adolescents, in which BFD is more clearly related to CV risk variables and the hip circumference is a 'low-risk' variable.

Detection of intrauterine growth retardation in twins using individualized growth assessment: I. Evaluation of growth outcome at birth.

The growth of 17 sets of twins was evaluated at 2 to 3 week intervals from 15 weeks to delivery by measurement of the head circumference, abdominal circumference, and thigh circumference, and estimation of weight. The birth characteristics of these twins were compared to those predicted by Rossavik growth models, derived from second-trimester ultrasound measurements, using the Growth Potential Realization Index (GPRI) and by comparison to population standards. Newborns were classified as normal or intrauterine growth retarded (IUGR) based on their Neonatal Growth Assessment Score, determined from GPRI values. All normal twins had birth weights that were appropriate-for-gestational-age and had few abnormal birth measurements. The birth weight differences between normal twin fetuses were all < 20%. IUGR twins were both small for gestational age (78%) and appropriate-for-gestational-age (22%), and all had 3 or 4 abnormal GPRI values. In only 40% of the cases was the birth-weight difference between a normal and an IUGR twin > 20%. No single anatomic parameter evaluated at birth adequately separated the normal twins from those with IUGR. These results indicate the need for multiple parameter Individualized Growth Assessment in the detection of IUGR in twins.

Detection of intrauterine growth retardation in twins using individualized growth assessment. II. Evaluation of third-trimester growth and prediction of growth outcome at birth.

Second- and third-trimester growth in 34 twin fetuses was evaluated with ultrasonography by measurement of five anatomic parameters. Rossavik growth models, derived from second-trimester measurements, were used to specify expected third-trimester growth curves. Actual measurements were compared to predicted measurements by calculation of the percent deviations. Growth outcome at birth [normal, intrauterine growth retardation (IUGR)] was determined from Neonatal Growth Assessment Scores. Growth in the second trimester was similar in normal and IUGR twins. In the third trimester, abnormal negative deviations were larger and more numerous in IUGR twins. However, there was considerable individual variability and normal twins also had abnormal negative deviations. In IUGR twins, the first appearance of an abnormal negative deviation was quite variable (range: 28.6 weeks to 35.1 weeks), as was the parameter to show such a deviation. Prediction of neonatal outcome was poor using individual anatomic parameters but improved considerably with use of all five parameters. However, some fetuses were misclassified when only the number of abnormal negative deviations was used. The Prenatal Growth Assessment Score (PGAS), determined by both the number and magnitude of abnormal negative deviations, predicted neonatal outcomes with a sensitivity of 100% and specificity of 100%. On average, PGAS values were abnormal 5 weeks before delivery. These results indicate that normal and IUGR twins can be separated, using third-trimester growth patterns, if multiple parameter Individualized Fetal Growth Assessment is employed.

Growth standards for anatomic measurements and growth rates derived from longitudinal studies of normal fetal growth.

A statistical procedure for deriving growth standards for anatomic measurements and their growth rates from longitudinal studies of fetal growth was evaluated using Rossavik growth models for the biparietal diameter (BPD), head circumference (HC), abdominal circumference (AC), and femur diaphysis length (FDL) determined in a previous study of normal fetal growth. For each anatomic parameter, the coefficients c and s of the model was used to define a set of growth curves that constituted the boundary growth curves of a region containing 95% of the growth curves of this data set. The set of boundary growth curves was used to specify the mean, lower limit, and upper limit values for the anatomic parameter and its growth rate at weekly intervals between 14 and 38 weeks, menstrual age. Comparison of these values to those determined from cross-sectional studies of fetal growth gave differences of -1.9% to 4.8% (SD: +/- 0.9 to +/- 2.6) for mean vs. predicted value of the anatomic measurements. For the lower limit, similar values were 0.4% to 13.8% (SD: +/- 1.7 to +/- 8.8); for the upper limit the values were 8.3% to 18.0% (SD: +/- 1.5 to +/- 7.0). Comparisons of HC growth rates determined using polynomial and Rossavik growth models gave values of -3.4% (SD: +/- 4.4) for mean vs. predicted value, -12.6% (SD: +/- 10.6) for the lower limit and 5.2% (SD: +/- 9.3) for the upper limit. The degree of agreement was similar for AC growth rates.(ABSTRACT TRUNCATED AT 250 WORDS)

Fetal body ratios in second trimester: a useful tool for identifying chromosomal abnormalities?

This study was carried out to determine if second trimester fetal body ratios are useful in detecting chromosomally abnormal fetuses. As a reference population, normative data for five fetal body ratios (femur length/biparietal diameter, biparietal diameter/fetal length, femur length/head circumference, head circumference/abdominal circumference, and femur length/abdominal circumference) were derived using regression analysis from a population of chromosomally normal fetuses (n = 1770) who underwent genetic amniocentesis at our institution between 14 and 21 menstrual weeks. During the same time period, 37 chromosomally abnormal fetuses were identified by amniocentesis. In comparing the two groups using the 10th and 90th percentiles as cutoffs between normal and abnormal, approximately 25% of chromosomally abnormal fetuses were identified, whereas approximately 20% of the normal fetuses were incorrectly classified as abnormal. Moreover, the use of 1.5 standard deviations above the mean for BPD/FL identified only 19% of Down syndrome fetuses. Our data, and those from a comprehensive review of the literature, suggest that the sensitivity of these ratios in detecting chromosomally abnormal fetuses is too low to recommend them for routine screening.

How accurate is second trimester fetal dating?

In this study, the Hadlock models for fetal dating using single and multiple parameters were tested retrospectively in 1770 chromosomally normal singleton fetuses in the second trimester (14 to 21 weeks of fetal development). The 95% confidence interval using measurements of the fetal head and femur individually was approximately +/- 1 week, which is comparable to the results of recently published dating models from other centers designed specifically for use during this time frame. The use of multiple-parameter models results in statistically significant improvement in prediction of age, in terms of both random error and maximum observed errors. We conclude that these models, developed for dating between 14 and 42 weeks of fetal development, provide highly accurate estimates of fetal age in the second trimester of pregnancy.

In utero analysis of fetal growth: a sonographic weight standard.

Regression analysis was used to develop an in utero fetal weight model from a population of 392 predominantly middle-class white patients with certain menstrual histories. There was a gradual increase in fetal weight from 35 g at 10 weeks to 3,619 g at 40 weeks, with uniform variance of +/- 12.7% (1 standard deviation) throughout gestation. When tested against the estimated weights of 1,771 chromosomally normal fetuses between 14 and 21 weeks, the mean percent difference was 0.8% and the average absolute percent error was 3.3%. When compared with actual delivery data for 163 fetuses in the group, the mean percent difference was 0.8% and the average absolute percent error was 1.1%. These data are compared with other prenatal weight curves obtained at ultrasound and with data from several large postnatal weight studies.

Exercise tolerance and alcohol intake. Blood pressure relation.

The relations of systolic and diastolic blood pressures to alcohol intake and exercise tolerance levels in 15,612 men and 3,855 women were investigated. Alcohol intake was assessed by questionnaire and stratified into seven levels for men and six for women according to the ounces of ethanol consumed per week. Exercise tolerance was determined by maximal treadmill exercise testing and was categorized into six age-specific by sex-specific levels. Both systolic and diastolic blood pressure were significantly related to both alcohol intake and exercise tolerance levels in both men and women. These relations, which were positive for alcohol and negative for exercise tolerance, remained after covariance adjustment for age, body mass index, and cigarette smoking. Alcohol intake was not significantly correlated with exercise tolerance. The relation of blood pressure to alcohol was not linear because the blood pressure of moderate consumers of alcohol tended to be slightly lower than that of nondrinkers. Higher blood pressure was found only in drinkers whose ethanol intake exceeded 9.5 ounces (approximately 285 ml or 19 drinks) per week. However, heavy drinkers in high exercise tolerance categories had no higher blood pressure than nondrinkers in low exercise tolerance groups. Exercise tolerance or physiological fitness appears to be important in quantifying the relation between alcohol intake and blood pressure and should be considered in describing this relation.

Neonatal growth assessment score: a new approach to the detection of intrauterine growth retardation in the newborn.

To develop an improved means for evaluation of the outcome of fetal growth in utero, a growth study of 37 fetuses at risk for intrauterine growth retardation was undertaken. Intrauterine growth was evaluated with ultrasonography at 2- to 3-week intervals beginning at 15 weeks' gestation. Measurements obtained in the second trimester were used to specify Rossavik growth models for various anatomic parameters. Values for weight, head circumference, abdominal circumference, and thigh circumference at birth were predicted with the use of these models. Growth potential realization index values for these four parameters were calculated from actual and predicted birth measurements. The growth potential realization index values were combined to form a neonatal growth assessment score. With the use of cluster analysis and neonatal growth assessment score values, the 37 infants were separated into 24 group I (neonatal growth assessment score values between 3.7 and 18.6) and 13 group II (neonatal growth assessment score values between 19.4 and 50.0) infants. In group I, 66.7% of the infants had no growth abnormalities; a single abnormality, usually borderline, was found in an additional 20.8% (total = 87.5%). In group II all infants had growth abnormalities (range, 2 to 6) although only 46.2% had birth weights below the 10th percentile. No differences in risk factors (except for the incidence of smoking) or second trimester growth were found in these two groups. The neonatal growth assessment score appears to be a sensitive indicator of third trimester growth retardation, is not affected by differences in growth potential, can separate normal infants from those with evidence of intrauterine growth retardation, and provides a quantitative assessment of growth problems in individual fetuses.